The nucleosomal acidic patch relieves auto-inhibition by the ISWI remodeler SNF2h.

Expression and purification of chromatin remodeling enzymes

SNF2h was purified from E. coli as described previously with minor modifications ( Leonard and Narlikar, 2015 ). DNA was precipitated following cell lysis by addition of 5% w/v polyethylenimine (P3143, Sigma-Aldrich, St. Louis, MO) pH 7.9 dropwise to a final concentration of 0.1% and clarified by centrifugation. Following cobalt affinity purification, the 6xHis tag was cleaved overnight with TEV protease and dialysed into SEC Buffer. TEV-cleaved SNF2h was then purified by anion exchange chromatography using a HiTrap Q column and size exclusion chromatography (GE Life Sciences, Pittsburgh, MA). SNF2h concentration was determined by SDS-PAGE with BSA protein standards and staining with SYPRO Red (Thermo Fisher, Waltham, MA). Human ACF was expressed and purified recombinantly from Sf9 insect cells by FLAG immunoaffinity purification as described previously with minor modifications ( Aalfs et al., 2001 ). SNF2h-FLAG and Acf1 were expressed separately via infection with baculovirus. Nuclear extracts from each construct were generated and mixed together at a 10:1 Acf1:SNF2h-FLAG volume ratio to ensure full assembly of the complex. This mixture was then bound to FLAG M2-affinity resin (Sigma-Aldrich, St. Louis, MO), washed with increasing KCl concentrations, and eluted with buffer with 100 mM KCl and 1 mg/mL FLAG Peptide. ACF concentration was determined by SDS-PAGE with BSA standards and based on the intensity of the Acf1 band. INO80 was purified by FLAG immunoprecipitation based on previously published methods ( Shen, 2004 ; Zhou et al., 2018 ). Briefly, S. cerevisae with endogenously flag-tagged INO80 was grown in YEPD at 30°C to saturation. Cells were pelleted by centrifugation for 10 min at 5000 rpm, resuspended with buffer H0.3 (25 mM HEPES, pH 7.5, 1 mM EDTA, pH 8.0, 10% glycerol, 0.02% NP-40, 0.3 M KCl), and pelleted again. Pelleted cells were then extruded through a 60 mL syringe into liquid nitrogen to create ‘noodles’. Cell ‘noodles’ were then lysed using a Tissue Lyser II (Qiagen, Hilden Germany) cooled in liquid nitrogen. Frozen lysate powder was resuspended in equal volume of H0.3 and spun in an SW28 rotor for 2 hr at 25,000 rpm at 4°C. Clarified lysate was mixed with equal volume buffer H0.3 and applied to FLAG M2-affinity resin (1 mL bead slurry per 40 mL of cleared lysate) equilibrated with H0.3 and incubated for 3 hr at 4°C. Resin was washed with 3 × 50 mL buffer H0.5 (H0.3 buffer except with 0.5 M KCl) followed by 3 × 10 mL washes with buffer H0.1 (0.1M KCl) and eluted with H0.1 supplemented with 1 mg/mL FLAG peptide. Eluate was concentrated, aliquoted, flash frozen in liquid nitrogen, and stored at −80°C. INO80 concentration was determined by SDS-PAGE with BSA standards, based on the intensity of the Ino80-flag band. Nucleosome labeling and reconstitution Recombinant Xenopus laevis histones were expressed and purified from E. coli as previously described ( Luger et al., 1999 ). Histone H2A E61A, E64A, D90A, D92A expression plasmid was a generous gift from the Tan lab at Penn State. Purified histone H2A E64R was provided by the Wolberger lab. Histone octamer was reconstituted as previously described ( Luger et al., 1999 ; Zhou and Narlikar, 2016 ), except for smFRET nucleosomes where a 2:1 unlabeled:labeled H3 mixture was used during octamer assembly to generate nucleosomes with mostly one H3 or neither H3 labeled. Histone H3 with a cysteine introduced at position 33 was labeled with either Cy3 (for smFRET) or Cy5 (for ensemble assays) prior to histone octamer assembly via cysteine-maleimide chemistry. Cy3-labeled (for ensemble assays) and Cyanine 5 SE-labeled and biotinylated DNAs (for smFRET) were generated by PCR with HPLC-purified, labeled primers (Cy5 primers: TriLink Biotechnologies, San Diego, CA; Cy3 and biotinylated primers: IDT, Coralville, IA) and purified by PAGE. The strong, synthetic 601 nucleosome positioning sequence ( Lowary and Widom, 1998 ) was used to assemble all nucleosomes in this study, with an arbitrary sequence for DNA flanking the 601 positioning sequence ( Figure 1—figure supplement 1 ). These DNAs were assembled with either wild-type or APM octamers by salt gradient dialysis and purified by glycerol gradient centrifugation ( Zhou and Narlikar, 2016 ).

Native gel remodeling assay

All remodeling reactions were performed under single turnover conditions (enzyme in excess of nucleosomes). Reactions with SNF2h were performed at 20°C with 20 nM nucleosomes, 12.5 mM HEPES pH 7.5, 2 mM Tris pH 7.5, 70 mM KCl, 5 mM ATP•MgCl 2 , 3 mM MgCl 2 , 0.02% NP40, and ~3%(v/v) glycerol. Reactions with ACF and INO80 were performed as above at 30°C and with minor changes in buffer composition (ACF: 10 nM nucleosomes, 12.5 mM HEPES pH 7.9, 2 mM Tris pH 7.5, 60 mM KCl, 2 mM ATP•MgCl 2 , 3 mM MgCl 2 , 0.02% NP40, 0.3 mg/mL FLAG peptide, and ~9% glycerol; INO80: 10 nM nucleosomes, 40 mM Tris pH 7.5, 60 mM KCl, 2 mM ATP•MgCl 2 , 1.1 mM MgCl 2 , 0.02% NP40, 0.5 mg/mL FLAG peptide, and 1% glycerol). Reactions were started with addition of enzyme and time points were quenched with excess ADP and plasmid DNA. Time points were then resolved by native PAGE (6% acrylamide, 0.5XTBE) and scanned on a Typhoon variable mode imager (GE Life Sciences, Pittsburgh, PA) by scanning for fluorescent labels. Gels were then quantified by densitometry using ImageJ. The fraction of nucleosomes end-positioned (i.e. unremodeled) at a given time point was determined by the ratio of fast-migrating nucleosomes to the total nucleosome intensity. This was fit to a single exponential decay using Prism 6 (GraphPad, La Jolla, CA) ( Equation 1 ), (1) y = ( y 0 − p ) e − k o b s t + p where y 0 is the initial fraction end-positioned, k obs is the observed rate constant, and p is the fraction end-positioned at plateau. Reactions in a given concentration series were fit constrained to a common y 0 and p . Concentration series were fit to a cooperative binding model ( Equation 2 ), (2) k o b s = k m a x X ( K m a p p ) h + X h where X is the concentration of SNF2h, h is the hill coefficient, K m app is the apparent K m , and k max is the saturating rate constant. Competition assays were performed as described above with varying concentrations of LANA peptide and fit to a single exponential decay. This was then fit to a simple competition binding model ( Equation 3 ), (3) k o b s = k 0 1 + X K 1 where k 0 is the rate constant without peptide, X is the concentration of peptide, and K I is the inhibition constant.

ATPase assay

ATPase reactions were performed under multiple turnover conditions (nucleosomes in excess of enzymes). Reactions were performed with 12.5 nM SNF2h, 12.5 mM HEPES pH 7.5, 2 mM Tris pH 7.5, 70 mM KCl, 7.5 µM ATP•MgCl 2 , 3 mM MgCl 2 , 0.02% NP40,~3%(v/v) glycerol, and trace amounts of γ- 32 P-ATP. Reactions were started with addition of enzyme, and 2.5 µL time points were quenched with an equal volume of 50 mM Tris pH 7.5, 3% SDS, and 100 mM EDTA. Inorganic phosphate was resolved from ATP on a PEI-cellulose TLC plate (Select Scientific) with 0.5 M LiCl/1M formic acid mobile phase. Plates were dried, exposed to a phosphorscreen overnight, and scanned on a Typhoon variable mode imager. Rate constants were determined by fitting a line through the first 10% of inorganic phosphate generated using Prism. Ensemble FRET remodeling assay Ensemble FRET remodeling assays were performed under the same conditions as gel remodeling assays. Reactions were initiated by addition of enzyme and then measured in a K2 fluorometer (ISS) equipped with a 550 nm short pass excitation filter and a 535 nm long pass emission filter. Reactions were excited at 515 nm and emission was measured at 665 nm. The resulting curves were fit to a two-phase exponential decay ( Equation 4 ), (4) y = ( p + ( y 0 − p ) ( f f a s t e − k f a s t t ) + ( 1 − f f a s t ) e − k s l o w t ) where f fast is the fraction in the fast phase and k fast and k slow are the apparent rate constants of the fast and slow phase respectively.

Crosslinking mass spectrometry

Crosslinking mass spectrometry samples were prepared by incubating 72 µg of mononucleosomes without flanking DNA at 9 µM final concentration with 24 µM SNF2h in buffer containing either ADP or ADP•BeF x (15 mM HEPES pH 7.5, 140 mM KCl, 0.5 mM ADP•Mg, 0.5 mM MgCl 2 ,±0.5 mM BeFx 1:5 BeCl 2 :NaF) for 10 min at 30°C. The crosslinking reaction with the LANA peptide was performed the same as the ADP•BeFx condition with 30 µM peptide added. The samples were then reacted with 1 mM EDC and 20 µM N-hydroxysulfosuccinimide (added as a 10x stock in water) for 60 min at room temperature. Crosslinking reactions were then quenched by adding 10 mM Tris pH 7.5 and samples were acetone precipitated and washed once with cold acetone. The pellet was resuspended in 8M Urea, 5 mM TCEP, 100 mM ammonium bicarbonate and heated at 56°C for 25 min, followed by alkylation with 10 mM iodoacetamide for 40 min at room temperature. The sample was diluted 5-fold with 100 mM ammonium bicarbonate and digested with 1:25 trypsin for 4 hr at 37°C followed by addition of a second aliquot of trypsin and overnight digestion. Crosslinked peptides were desalted using 100 µl OMIX C18 tips (Agilent), fractionated by size-exclusion chromatography (SEC), and analyzed by LC-MS similarly to a previously described method ( Zhou et al., 2017 ). Briefly, trypsin digests were acidified to 0.2% TFA, desalted, and run over a Superdex Peptide PC 3.2/300 s column (GE Healthcare). SEC fractions eluting between 0.9 ml and 1.4 ml were dried and resuspended in 0.1% formic acid for LC-MS. Each fraction was separated over a 15 cm x 75 μm ID PepMap C18 column (Thermo) using a NanoAcquity UPLC system (Waters) and analyzed by a Fusion Lumos mass spectrometer (Thermo). Precursor ions were measured from 375 to 1500 m/z in the Orbitrap analyzer (resolution: 120,000; AGC: 4.0e5). Ions charged 3+ to 8+ were isolated in the quadrupole (selection window: 1.6 m/z units; dynamic exclusion window: 30 s; MIPS Peptide filter enabled), fragmented by HCD (Normalized Collision Energy: 28%) and measured in the Orbitrap (resolution: 30,000; AGC; 5.0e4). The cycle time was set to 3 s. Peaklists were generated using PAVA (UCSF) and searched for crosslinked peptides with Protein Prospector 5.19.22 ( Trnka et al., 2014 ) against a target database containing human SNF2h plus the four core histone sequences from X. laevis concatenated with a decoy database containing 10 randomized copies of each target sequence (total database size: 55 sequences). Loss of the initiator methionine and carbamidomethylation of cysteine. Methionine oxidation, peptide N-terminal glutamine to pyroglutamate formation, acetylation at the protein N-terminus, and mis-annotation of the monoisotopic peak (1 Da neutral loss) were treated as variable modifications. EDC was designated as a heterobifunctional crosslinking reagent with specificity of aspartate, glutamate, and the protein C-terminus on one side and lysine and the protein N-terminus on the other with a bridge mass corresponding to loss of H 2 O. A mass modification range of 400–5000 Da was specified on these residues and 85 product ion peaks from the peaklist were used in the search. Precursor and product ion tolerances were 8 and 25 ppm respectively. Crosslinked spectral matches (CSMs) were initially classified as in ( Zhou et al., 2017 ). The dataset was then aggregated into unique crosslinked residue-pair level data with a corresponding spectral count value. Due to the prevalence of multiple, closely spaced Asp and Glu residues in a typical tryptic peptide, site-localization of EDC crosslinks is more challenging than with homobifunctional lysine-directed reagents. To address this, when the site-localization was judged to be ambiguous, all possible residue-pairs were kept with an annotation noting the ambiguity. When calculating spectral counts, fractional spectral counts were assigned to these ambiguous site localizations so that a given spectrum was awarded exactly one spectral count. For instance, a product ion spectrum matching equally well to both K91.H4-D65.H2B or K91.H4-E68.H2B contributes 0.5 towards the spectral counts of each residue-pair. Decoy CSMs were retained throughout this aggregation and spectral counting process. A linear SVM model, built on five features of the Protein Prospector search output (score difference, % of product ion signals matched, precursor charge, rank of peptide 1, and rank of peptide 2) was constructed to sort crosslinked residue pairs into decoy and target classes. Crosslinked residue-pairs with an SVM score greater than 1, score difference greater than 5, and at least one spectral count are reported. The final residue-pair level data set is reported at specificity of 99.7% corresponding to 0.05% FDR. The number of unique crosslinks in the ADP condition was 974, while 1470 crosslinks were unique to the ADP•BeF x condition and 707 crosslinks were common to both conditions. To determine which protein domains are involved in SNF2h mediated nucleosome sliding, residue level crosslink spectral counts were aggregated into domain level counts. Each domain pair was assigned a minimum spectral count of 1 to avoid dividing by zero, and the log2 ratio of spectral counts for each domain pair in the ADP•BeFx condition to the ADP condition were calculated. Domain pairs with a Log2 ratio of exactly 0 were treated as NA values. The remaining data were normalized such that the median value was set at 0. Hence, most domain-domain interactions were assumed to not change substantially between conditions. Histone protein sequences were taken from Xenopus Laevis Uniprot Entries (without the initiator methionine) and domains were defined as follows: H2A N-term tail: 1–16, H2A: 17–43, H2A Acidic Patch: 44–100, H2A C-term tail: 101–129; H2B tail: 1–34, H2B: 35–99, H2B Acidic Patch: 100–122; H3 tail: 1–44, H3: 45–135; H4 tail: 25–102. The sequence of SNF2h was identical to the Human entry in Uniprot ( O60264 ) with an additional two amino acids at the N-terminus to match the construct used. All residue numbers are therefore shifted from the Uniprot entry by two aa. SNF2h domains were defined as follows: Snf2h1: 1–83, AutoN: 84–160, Snf2h2: 161–183, RecA1: 184–402, RecA2: 403–641, NegC: 642–703, Snf2h4: 704–736, HAND: 737–839, SANT: 840–894, SLIDE: 895–1013, Snf2h5: 1014–1054. Annotated Mass Spectra are available using MS-Viewer at: http://msviewer.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msviewer The ADP data set is accessed with search key: 2x0kr2kzq1 The ADP•BeFx data set is accessed with search key: fjamygr8pl The ADP•BeFx in the presence of LANA peptide is accessed with search key: c5o2mcxwum. Raw mass spectrometry data is available in the MassIVE repository at UCSD with accession key: MSV000082136 Single molecule FRET measurements smFRET experiments were performed as previously described in ( Zhou et al., 2018 ) with modifications to the reaction buffers as described below.

Sample preparation and imaging

Briefly, as in ( Zhou et al., 2018 ), quartz slides were PEGylated and then incubated with neutravidin (A2666, ThermoFisher Scientific, Waltham, MA) to mediate attachment of biotinylated nucleosomes. After removal of unbound neutravidin, biotinylated nucleosomes at 12.5 pM in a modified Wash Buffer (12 mM HEPES-KOH, pH 7.5 at 22°C, 60 mM KCl, 1.4 mM MgCl 2 , 10% glycerol, 0.1 mM EDTA, 0.02% Igepal, 1% [w/v] glucose, and 0.1 mg/mL acetylated BSA) were added and incubated for 10 min. Unbound nucleosomes were removed by washing with Wash Buffer. All incubations and experiments were performed at 20°C. Nucleosomes were imaged on a custom-built prism-based TIRF setup. Immediately prior to data acquisition, the sample chamber was flushed with a modified imaging buffer (53 mM HEPES-KOH, pH 7.5 at 22°C, 9.1 mM Tris-acetate, pH 7.5 at 22°C [contributed by the Trolox], 63 mM KCl, 1.41 mM MgCl 2 , 10% glycerol, 0.1 mM EDTA, 0.02% Igepal, 1% [w/v] glucose, 0.1 mg/mL acetylated BSA, 2 mM Trolox [Sigma 238813, made as an 11 mM stock in Tris-acetate, pH'd to 7.5 with 1 M NaOH, and stored at 4°C], 0.03 mM β-mercaptoethanol, 2 U/μL catalase [=0.2 mg/mL; Sigma E3289], and 0.08 U/μL glucose oxidase [0.8 mg/mL, Sigma G2133; made with the catalase as a 100x stock in SPB, and stored at 4°C for not more than one week]). Images were collected using Micro-Manager ( www.micro-manager.org , San Francisco, CA) ( Edelstein et al., 2010 ) at 7.4 Hz, with an exposure time of 100 ms. To start each reaction, saturating enzyme (400 nM for WT nucleosomes; 2 μM for E64R nucleosomes) and saturating ATP (1 mM) in 300 μL imaging buffer were added via an automated syringe pump (J-KEM Scientific, St. Louis, MO).

Data analysis

The number of remodeling events per smFRET experiment was roughly an order of magnitude lower with E64R nucleosomes compared to WT nucleosomes, necessitating significantly larger data sets for the E64R nucleosomes compared to WT. To streamline the data analysis process with these data sets, as well as to improve the overall quality of the data, we made use of the custom in-house software we have developed for smFRET image analysis, called Traces, available for download at https://github.com/stephlj/Traces ( Zhou et al., 2018 ; Johnson et al., 2018 ; copy archived at https://github.com/elifesciences-publications/Traces ). In addition, the long pauses exhibited by SNF2h remodeling E64R nucleosomes are subject to an increased number of artifacts, such as dye blinking or slight fluctuations in the noise, which can complicate quantification of smFRET trajectories. We therefore used the python-based HMM library pyhsmm ( https://github.com/mattjj/pyhsmm ), which we adapted for the analysis of smFRET data as part of the Traces package, to quantify pause durations. This particular HMM package, which fits a discrete state HMM to each trajectory generated by Traces, enables a reduction in the likelihood of the HMM identifying artifacts as real transitions, and also reduces analysis time. As described in Figure 3—figure supplement 1 , and consistent with previous smFRET studies with nucleosomes ( Blosser et al., 2009 ; Deindl et al., 2013 ; Hwang et al., 2014 ), we observe two predominant clusters of FRET values, at 0.57 and 0.95 FRET, in the absence of remodeler. These FRET states correspond to two of the four populations of nucleosomes that result from mixing unlabeled H3 with Cy3-labeled H3 during octamer formation: some nucleosomes will have a Cy3 label on the H3 proximal to the Cy5-labeled DNA end, resulting in the higher FRET state, and some will have a Cy3 label on the H3 distal to the Cy5-labeled DNA end, resulting in the mid-FRET state. There will also be a population of nucleosomes with both H3 histones unlabeled, which show no FRET; and a population with both copies of H3 labeled, which are distinguishable by two-step photobleaching of the Cy3 dye, and are excluded from all analyses (i.e. both calibration curve data ( Zhou et al., 2018 ) and remodeling data). We also excluded any trajectories to which pyhsmm fit an initial FRET value lower than 0.775 FRET, since nucleosomes with distally labeled H3's do not provide as great a dynamic range for monitoring nucleosome remodeling. We excluded any part of a trajectory including and subsequent to backtracking events (where the nucleosome was moved away from the center, that is, where FRET increased instead of decreasing). Each data set (e.g., E64R/SNF2h) consists of at least 100 trajectories collated from at least 3 (typically 5 to 7) different experiments. All errors were bootstrapped over trajectories (that is, for each data set, the ≥100 trajectories were resampled with replacement, and reported values, such as the means of each pause duration, were recalculated for each bootstrapped sample). Reported errors are standard deviations of the bootstrapped values. A similar procedure was used to obtain the errors on the cdfs in Figure 3E and Figure 3—figure supplement 3 ; the shaded regions represent ±the standard deviation of the set of bootstrapped cdfs.

Additional files 10.7554/eLife.35322.028 Supplementary file 1. Crosslinked residue pairs identified from EDC treatment of SNF2h-nucleosomes in the presence of: ADP, ADP-BeFx, or ADP-BeFx and LANA peptide. 10.7554/eLife.35322.029 Transparent reporting form Data availability Relevant source data is provided in the main and supplemental figures. Crosslinked residue pair identification along with number of spectral counts per identification are reported in Supplementary file 1, as well as in a web resource with links to annotated product ion spectra (see Experimental Methods). Raw mass spectrometry files are available on the Massive server (UCSD). Code used for the analysis of smFRET data can be found at the following link, which is also found in the main text. https://github.com/stephlj/Traces The following datasets were generated: Trnka MJ 2018 Annotated product ion peaklists - ADP crosslinked dataset http://msviewer.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msviewer Publicly available at the UCSF MS-Viewer (search key 2x0kr2kzq1) Trnka MJ 2018 Annotated product ion peaklists - ADP•BeFx crosslinked dataset http://msviewer.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msviewer Publicly available at the UCSF MS-Viewer (search key fjamygr8pl) Trnka MJ 2018 Unprocessed Mass Spectrometry Files https://massive.ucsd.edu/ProteoSAFe/dataset.jsp?task=4b2c8ce69fb44e89890b7617728d19e6 Publicly available at MassIVE (identifier MSV000082136) Trnka MJ 2018 Annotated product ion peaklists - ADP•BeFx with LANA peptide crosslinked dataset http://msviewer.ucsf.edu/prospector/cgi-bin/msform.cgi?form=msviewer Publicly available at the UCSF MS-Viewer (search key c5o2mcxwum)