A Method for Genetically Installing Site-Specific Acetylation in Recombinant Histones Defines the Effects of H3 K56 Acetylation

Summary Lysine acetylation of histones defines the epigenetic status of human embryonic stem cells and orchestrates DNA replication, chromosome condensation, transcription, telomeric silencing, and DNA repair. A detailed mechanistic explanation of these phenomena is impeded by the limited availability of homogeneously acetylated histones. We report a general method for the production of homogeneously and site-specifically acetylated recombinant histones by genetically encoding acetyl-lysine. We reconstitute histone octamers, nucleosomes, and nucleosomal arrays bearing defined acetylated lysine residues. With these designer nucleosomes, we demonstrate that, in contrast to the prevailing dogma, acetylation of H3 K56 does not directly affect the compaction of chromatin and has modest effects on remodeling by SWI/SNF and RSC. Single-molecule FRET experiments reveal that H3 K56 acetylation increases DNA breathing 7-fold. Our results provide a molecular and mechanistic underpinning for cellular phenomena that have been linked with K56 acetylation.

Protein total mass was determined on an LCT time-of-flight mass spectrometer with electrospray ionization (ESI). (Micromass). Proteins were rebuffered to 20 mM (NH 4 )HCO 3 pH 7.5 and diluted 1:100 into 50% methanol, 1% formic acid. Samples were infused into the ESI source at 10 ml min -1 , using a Harvard Model 22 infusion pump (Harvard Apparatus) and calibration performed in positive ion mode using horse heart myoglobin. 60-80 scans were acquired and added to yield the mass spectra. Molecular masses were obtained by deconvoluting multiply charged protein mass spectra using MassLynx version 4.1 (Micromass). Theoretical molecular masses of wild-type proteins were calculated using Protparam (http://us.expasy.org/tools/protparam.html), and theoretical masses for unnatural amino acid containing proteins adjusted manually. Where indicated protein total mass and acetylation position sequencing was performed using a top down approach, in these cases in-source decay (ISD) spectra were acquired in reflectron mode on an Ultraflex III TOF/TOF mass spectrometer (Bruker Daltonics, Bremen, Germany) using a 2,5-dihydroxy benzoic acid matrix.

Histone octamer reconstitution
Lyophilized histones were dissolved at an equivalent of 1 mg H2A per ml in unfolding buffer (7 M guanidinium chloride in 20 mM Tris, pH 7.4, 10 mM DTT) and mixed in stoichiometric amounts (Luger et al., 1999). A 2 ml reaction was incubated for 3 h at room temperature with gentle agitation and dialysed against three times 250 ml refolding buffer (2 M NaCl, 10 mM Tris pH 7.4, 1 mM EDTA, 5 mM βmercaptoethanol) at 4ºC. Precipitates were removed by centrifugation (5 min, 25000 g, 2ºC) and filtered using a SpinX column. Octamers were then separated by gel filtration using a Superdex200 column equilibrated with refolding buffer.

Labelling of H2A K119C with maleimide-Cy5
The K119C mutation was introduced into pET3 H2A by Quikchange and H2A K119C was expressed and purified following published procedures (Luger et al., 1999). The protein was rebuffered to degassed PBS containing 1 mM TCEP using a PD10 column. In a 1 ml reaction 2 mg of the protein were reacted with 400 µg maleimide-Cy5 for 18 h at 4ºC. The reaction was then dialysed against two times 500 ml 5 mM β-mercaptoethanol over night at 4ºC and lyophilized. Analysis by ESI-TOF MS showed that the reaction had gone to completion (See supplementary information).
Histone octamers containing unacetylated H3 or H3 K56Ac and Cy5-labelled H2A were reconstituted as described above. Octamers and 147 bp Cy3-DNA were mixed in high-salt buffer (2 M NaCl, 10 mM Tris-HCl (pH 7.4), 1 mM EDTA, 5 mM βmercaptoethanol) and nucleosome core particles assembled by a continuous dialysis method in which the NaCl concentration was reduced from 2.0 M to 10 mM over a 16 hour period at 4°C. The stoichiometry of histone octamer binding was assessed by gel mobility-shift assays in 0.8% (w/v) agarose gels imaged with a Typhoon Imager.
Fluorescence experiments were carried out at room temperature (~23°C) on a Tecan safire 2 spectrophotometer. Nucleosome samples were excited at 515 nm and emission spectra were collected from 535-750 nm. Emission wavelength maxima were observed at 565 nm for Cy3 and 670 nm for Cy5. Samples were incubated for at least 5 minutes at each salt concentration prior to each reading, as it has been previously demonstrated that longer incubation does not lead to any further change in emission intensity (Park et al., 2004), indicating that an equilibrium has been achieved within 5 min. All samples contained a final concentration of ~8 nM nucleosome core particles.
Relative fluorescence intensity was calculated from FRET donor intensity/ FRET acceptor intensity and data were normalized using the upper and lower plateau values as baselines.

Production of DNA arrays for compaction
To produce and purify the DNA arrays, E. coli DH5α containing a pUC18 vector with the DNA array insert was grown overnight in 1 L of 2×TY (37°C, 210 rpm). For blunt-ended release, multimer arrays (12 kbp) were excised by digestion with EcoRV.
The vector was digested into smaller products (<1 kbp) using HaeII and DraI. The array DNA was separated from the fragments by selective polyethylene glycol (PEG) precipitation of long DNA fragments using 5-8% PEG 6000 in 0.5 M NaCl. The purified array DNA was phenol/chloroform extracted, ethanol precipitated, and the DNA pellets were re-suspended in 2 M NaCl, 10 mM TEA and 1 mM EDTA.
Competitor DNA (crDNA) was obtained from chicken erythrocyte nuclei. Mononucleosomes with approximately 147 bp of mixed sequence DNA were obtained by limited micrococcal nuclease digest of long chicken chromatin. Phenol/chloroform extraction removed bound histones.

Reconstitution of nucleosome arrays
Nucleosome arrays were reconstituted at 25 µg/ml DNA using our in vitro reconstitution method (Huynh et al., 2005). The molar input ratio of histone octamer required to obtain saturation was empirically determined. For compaction studies, the linker histone (H5) was added to the reconstitution in increasing concentrations.

Sedimentation velocity analysis data
Sedimentation velocity analysis data were obtained using a Beckman XL-A analytical ultracentrifuge equipped with scanner optics. Optical density was measured at 260 nm with an initial absorbance between 0.5 and 1.2. Sedimentation analysis was carried out for 2 h at 5°C at speeds between 18,000 r.p.m. in 12 mm double-sector cells and a Beckman AN60 rotor. Prior to analysis, samples and blanking buffer were placed in cells to settle for approximately one hour as this dramatically improved reproducibility of results. Sedimentation coefficients were determined using the timederivative method described by Stafford (Stafford, 1992), using John Philo's Dcdt+ data analysis program (version 2.05) (Philo, 2006). Sedimentation coefficients were corrected to S 20,w . Partial specific volumes were calculated for all sample assuming values of 0.725 and 0.55 for protein content and DNA content respectively. Partial specific volumes are thus adjusted to account for linker histone content. Solvent viscosity and solvent density were corrected according to buffer composition.

Purification of remodeling complexes
Yeast strains TAP tagged for RSC (Saha et al., 2002) and SWI/SNF (Chandy et al., 2006) were purified as described previously (Ferreira et al., 2007). The SWI/SNF used for testing H3 K56 acetylated nucleosomes was a kind gift from Salma Mahmood and was purified essentially as described (Ferreira et al., 2007) but with the following changes: 6 L of cells were grown in 1×yeast extract, peptone, adenine, Dglucose. The cells were disrupted using 0.5 mm glass beads in a Bead Beater (Biospec Products Incorporated) using 10 pulses of 30 s ON, 1 min OFF. SWI/SNF wash and storage buffers contained 150 mM NaCl.

Mononucleosome repositioning assays
Nucleosomes were assembled onto DNA fragments described with the nomenclature  Phosphoimager FLA-5100 (Fujifilm, Japan). Gel band intensities were quantitated using AIDA software (Raytest, Germany) and the remodeller repositioning at each time point calculated from the intensity of the sum of all end positions relative to the sum of the major initial and all end positions. The initial rate was calculated as previously described (Ferreira et al., 2007). Each initial rate was repeated at least three times using chromatin prepared in separate assembly reactions.

Dimer exchange assays
Histone H2A T10C was fluorescently labelled by a Cy5 mono maleimide dye (GE Healthcare). Donor nucleosomes were produced by assembly of tetramers and Cy5 labelled dimers onto Cy3 labelled 54A18 DNA fragments. To measure nucleosome assembly efficiency 2 pmol of each assembly reaction was resolved by native PAGE and the assembly quantified by measuring the summed intensity of all nucleosome bands relative to 1 pmol of Cy3 labelled 54A18 DNA. Each 10 μl reaction contained 0.25 pmol of donor nucleosome, 0.75 pmol (3 fold excess) wild-type tetrasome acceptor assembled on 0W0 DNA fragments, 50 mM NaCl, 50 mM Tris pH 7.5, 3 mM MgCl 2 , 1 mM ATP and the quantity of remodeller specified in figure 6.
Reactions were incubated in an Eppendorf mastercycler with heated lid at 30°C for the specified times. Reactions were terminated by transfer to ice and the addition of 500 ng of HindIII-digested bacteriophage lambda competitor DNA (Promega, USA) and 5% (w/v) sucrose. Samples were resolved on a native PAGE gel and the