Genome-wide reconstitution of chromatin transactions reveals that RSC preferentially disrupts H2AZ-containing nucleosomes.

Chromatin transactions are typically studied in vivo, or in vitro using artificial chromatin lacking the epigenetic complexity of the natural material. Attempting to bridge the gap between these approaches, we established a system for isolating the yeast genome as a library of mononucleosomes harboring the natural epigenetic signature, suitable for biochemical manipulation. Combined with deep sequencing, this library was used to investigate the stability of individual nucleosomes and, as proof of principle, the nucleosome preference of the chromatin remodeling complex, RSC. This approach uncovered a distinct preference of RSC for nucleosomes derived from regions with a high density of histone variant H2AZ, and this preference is indeed markedly diminished using nucleosomes from cells lacking H2AZ. The preference for H2AZ remodeling/nucleosome ejection can also be reconstituted with recombinant nucleosome arrays. Together, our data indicate that, despite being separated from their genomic context, individual nucleosomes can retain their original identity as promoter- or transcription start site (TSS)-nucleosomes. Besides shedding new light on substrate preference of the chromatin remodeler RSC, the simple experimental system outlined here should be generally applicable to the study of chromatin transactions.

Isolation of an activator-dependent, promoter-specific chromatin remodeling factor.

Repressed PHO5 gene chromatin, isolated from yeast in the native state, was remodeled by yeast extract in a gene activator-dependent, ATP-dependent manner. The product of the reaction bore the hallmark of the process in vivo, the selective removal of promoter nucleosomes, without effect on open reading frame nucleosomes. Fractionation of the extract identified a single protein, chromodomain helicase DNA binding protein 1 (Chd1), capable of the remodeling activity. Deletion of the CHD1 gene in an isw1Δ pho80Δ strain abolished PHO5 gene expression, demonstrating the relevance of the remodeling reaction in vitro to the process in vivo.

Structure-guided engineering of xylitol dehydrogenase cosubstrate specificity.

Xylitol dehydrogenase (XDH) is one of several enzymes responsible for assimilating xylose into eukaryotic metabolism and is useful for fermentation of xylose contained in agricultural byproducts to produce ethanol. For efficient xylose utilization at high flux rates, cosubstrates should be recycled between the NAD+-specific XDH and the NADPH-preferring xylose reductase, another enzyme in the pathway. To understand and alter the cosubstrate specificity of XDH, we determined the crystal structure of the Gluconobacter oxydans holoenzyme to 1.9 angstroms resolution. The structure reveals that NAD+ specificity is largely conferred by Asp38, which interacts with the hydroxyls of the adenosine ribose. Met39 stacked under the purine ring and was also located near the 2' hydroxyl. Based on the location of these residues and on sequence alignments with related enzymes of various cosubstrate specificities, we constructed a double mutant (D38S/M39R) that was able to exclusively use NADP+, with no loss of activity.

Structural and catalytic diversity in the two family 11 aldo-keto reductases.

Aldo-keto reductases (AKRs) are a large superfamily of NAD(P)H-dependent enzymes that function in a wide range of biological processes. The structures of two enzymes from the previously uncharacterized family 11 (AKR11A and AKR11B), the products of the iolS and yhdN genes of Bacillus subtilis have been determined. AKR11B appears to be a relatively conventional member of the superfamily with respect to structural and biochemical properties. It is an efficient enzyme, specific for NADPH and possesses a catalytic triad typical for AKRs. AKR11A exhibits catalytic divergence from the other members of the superfamily and, surprisingly, AKR11B, the most closely related aldo-keto reductase in sequence. Although both have conserved catalytic residues consisting of an acidic tyrosine, a lysine and an aspartate, a water molecule interrupts this triad in cofactor-bound AKR11A by inserting between the lysine and tyrosine side-chains. This results in a unique architecture for an AKR active site with scant catalytic power. In addition, the absence of a bulky tryptophan side-chain in AKR11A allows an unconventional conformation of the bound NADP+ cosubstrate, raising the possibility that it donates the 4-pro-S hydride rather than the 4-pro-R hydride seen in most other AKRs. Based upon the architecture of the active site and the resulting reaction velocities, it therefore appears that functioning as an efficient oxido-reductase is probably not the primary role of AKR11A. A comparison of the apo and holo forms of AKR11A demonstrates that the cosubstrate does not play the dramatic role in active site assembly seen in other superfamily members.

Retention of the Native Epigenome in Purified Mammalian Chromatin.

A protocol is presented for the isolation of native mammalian chromatin as fibers of 25-250 nucleosomes under conditions that preserve the natural epigenetic signature. The material is composed almost exclusively of histones and DNA and conforms to the structure expected by electron microscopy. All sequences probed for were retained, indicating that the material is representative of the majority of the genome. DNA methylation marks and histone marks resembled the patterns observed in vivo. Importantly, nucleosome positions also remained largely unchanged, except on CpG islands, where nucleosomes were found to be unstable. The technical challenges of reconstituting biochemical reactions with native mammalian chromatin are discussed.

Head module control of mediator interactions.

Yeast Mediator proteins interacting with Med17(Srb4) have been expressed at a high level with the use of recombinant baculoviruses and recovered in homogeneous form as a seven subunit, 223 kDa complex. Electron microscopy and single-particle analysis identify this complex as the Mediator head module. The recombinant head module complements "headless" Mediator for the initiation of transcription in vitro. The module interacts with an RNA polymerase II-TFIIF complex, but not with the polymerase or TFIIF alone. This interaction is lost in the presence of a DNA template and associated RNA transcript, recapitulating the release of Mediator that occurs upon the initiation of transcription. Disruption of the head module in a temperature-sensitive mutant in vivo leads to the release of middle and tail modules from a transcriptionally active promoter. The head module evidently controls Mediator-RNA polymerase II and Mediator-promoter interactions.