Structural Basis of Dot1L Stimulation by Histone H2B Lysine 120 Ubiquitination.

CONTACT FOR REAGENT AND RESOURCE SHARING

Further inquiries and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Karim-Jean Armache ( karim-jean.armache@nyulangone.org ). METHOD DETAILS Expression and purification of Dot1L and mutants Plasmid DNA of pET28-MHL-Dot1L (1-420) was purchased from Addgene (#40736), and respective Dot1L mutants were generated in lab using Q5 mutagenesis kit (NEB). Expression and purification steps for Dot1L and its mutants were the same. Dot1L plasmid DNA was transformed into ONESHOT™BL21(DE3) (ThermoFisher) competent cells and grown in 2xYT-Kan media. Dot1L was expressed as soluble protein by induction with 0.5 mM IPTG for 3 hours at 37°C upon the culture reaching OD600 = 0.4-0.6. Cells were harvested (Sorvall LYNX6000) and lysed (AvestinEmulsiflexC3), and the protein was purified through Ni-NTA agarose beads (Qiagen) (Lysis Buffer: 500 mM NaCl, 50 mM Tris pH 8.0, 5 mM Imidazole, 5% Glycerol, 2mM β-Mercaptoethanol (BME), 1× Protease Inhibitor (cOmplete EDTA free, Roche) / Elution Buffer: 500 mM NaCl, 50 mM Tris pH 8.0, 300 mM Imidazole, 5% Glycerol, 2 mM BME). Eluted Dot1L protein was digested with TEV protease (NEB) while dialyzing in digestion buffer (Digestion Buffer: 75 mM NaCl, 20 mM Tris pH 8.0, 5% Glycerol, 2 mM BME) overnight at 40°C to cleave off Polyhistidine-tag. After tag was cleaved off, sample was purified through HiTrap SP HP (GE Healthcare) liquid chromatography column (Buffer A: 75 mM NaCl, 25 mM HEPES pH 7.5, 5% glycerol, 2 mM BME / Buffer B: 1 M NaCl, 25 mM HEPES pH 7.5, 5% glycerol, 2 mM BME). Selected fractions were further purified through HiLoad Superdex 200 16/600 (GE Healthcare) size-exclusion liquid chromatography column (S200 Buffer: 150 mM NaCl, 10 mM HEPES pH 7.5, 2 mM DTT). Purified Dot1L protein was then concentrated, flash frozen in liquid nitrogen and stored in −80°C for future use.

Expression and purification of yeast Dot1 Plasmid

DNA of pET15b-His-3C-Dot1 (158-582) was generated using Gibson Assembly (NEB).

Yeast Dot1 plasmid

DNA was transformed into ONESHOT™BL21(DE3) (ThermoFisher) competent cells and grown in 2xYT-Amp media. Yeast Dot1 was expressed as soluble protein by inducing with 0.1mM IPTG for 16 hours at 18°C upon the culture reaching OD600 = 0.4-0.6. Cells were harvested (Sorvall LYNX6000) and lysed (AvestinEmulsiflexC3), and the protein was purified through Ni-NTA agarose beads (Qiagen) (Lysis Buffer: 500 mM NaCl, 20 mM HEPES pH 7.5, 20 mM Imidazole, 1 mM BME, 1× Protease Inhibitor / Elution Buffer: 500 mM NaCl, 20 mM HEPES pH 7.5, 300 mM Imidazole, 1 mM BME). Eluted yeast Dot1 protein was digested with 3C PreScission Protease (GE Healthcare) while dialyzed in digestion buffer (Digestion Buffer: 100 mM NaCl, 20 mM HEPES pH 7.5, 1 mM DTT) overnight at 4°. After which, sample was purified through HiTrap SP HP (GE Healthcare) liquid chromatography column (Buffer A: 100 mM NaCl, 20 mM HEPES pH 7.5, 1 mM DTT / Buffer B: 1 M NaCl, 20 mM HEPES pH 7.5, 1 mM DTT). Selected fractions were further purified through HiLoad Superdex 200 (GE Healthcare) size-exclusion liquid chromatography column (S200 Buffer: 150 mM NaCl, 10 mM Tris pH 7.5, 5 mM DTT). Purified yeast Dot1 protein was then concentrated, flash frozen in liquid nitrogen and stored in −80°C for future use.

Expression and purification of Ubiquitin Plasmid DNA of pET-His-UB

G76C was a generous gift from Dr. Tingting Yao. Ubiquitin plasmid DNA was transformed into SoluBL21™ (amsbio) competent cells and grown in 2xYT-Amp Media. Ubiquitin was expressed as soluble protein by inducing with 0.4 mM IPTG for 4 hours at 37°C upon the culture reaching OD600 = 0.4-0.6. Bacteria cells were harvested and lysed (AvestinEmulsiflexC3). Protein was then purified through Ni-NTA agarose beads (Qiagen) (Lysis Buffer: 300 mM NaCl, 50 mM Tris pH 8.0, 10 mM Imidazole, 5 mM BME, 1× Protease Inhibitor / Elution Buffer: 300 mM NaCl, 50 mM Tris pH 8.0, 300 mM Imidazole, 5 mM BME) followed by HiTrap Q HP (GE Healthcare) liquid chromatography column (Buffer A: 50 mM NaCl, 20 mM Tris pH 8.0, 0.2 mM EDTA, 10 mM BME / Buffer B: 1 M NaCl, 20 mM Tris pH 8.0, 0.2 mM EDTA, 10 mM BME). Purified ubiquitin was then dialyzed against water supplemented with 1 mM acetic acid followed by flash freezing in liquid nitrogen and lyophilized using VIRTISSentry. Purification of Widom 601 DNA A plasmid containing 8 copies of the Widom 601 nucleosome positioning sequence, each flanked by the EcoRV restriction enzyme cutting site ( Armache et al., 2011 ), was transformed into DH5α™ (ThermoFisher) competent cells and grown in 2xYT-Amp media overnight. The 601 DNA fragment was excised using EcoRV and purified using previously published protocols ( Dyer et al., 2004 ).

Expression and purification of wild-type Xenopus

Histones and Histone mutant H2BK120C Plasmids containing wild-type Xenopus histones were a generous gift from Dr. Karolin Luger, and mutant histone H2B-K120C was generated using the Q5 mutagenesis kit (NEB). Briefly, each histone was expressed in Rosetta (DE3) cells (Novagen), extracted from inclusion bodies, and purified sequentially by size exclusion and anion chromatography using previously published protocols ( Dyer et al., 2004 ). Purified histones were freeze dried using a Sentry lyophilizer (VirTis). Ubiquitination of Histone H2BK120C We followed previously published protocol ( Long et al., 2014 ). Briefly, lyophilized ubiquitin and histone H2BK120C were re-suspended (Re-suspension buffer: 10 mM Acetic acid, 7M Urea-Deionized) and mixed in the ratio of 2:1. Sodium Tetraborate, Urea and TCEP were added to achieve final concentration of 50 mM, 6M and 5 mM, respectively. Final solution was incubated at room-temperature for 30 minutes. Then, an amount of crosslinker (Di-Chloracetone diluted in Di-Methylformamide) equal to one-half molar ratio of total sulfhydryl groups was added to the solution and incubated on ice for additional 30 minutes. Reaction was ended by addition of BME to final concentration of 5 mM. Solution was then diluted 10 times with Denaturing Binding Buffer (Denaturing Binding Buffer: 50 mM Sodium phosphate (NaPi), 50 mM Tris pH 8.0, 300 mM NaCl, 6 M Urea, 10 mM Imidazole, 5 mM BME) and purified through Ni-NTA agarose beads (Qiagen) (Denaturing Elution Buffer: 50 mM NaPi, 50 mM Tris pH 8.0, 300 mM NaCl, 6 M Urea, 250 mM Imidazole, 5 mM BME). Purified Ubiquitinated-H2BK120C was dialyzed into water supplemented with 1 mM BME and lyophilized using VIRTISSentry. Reconstitution of nucleosomes Unmodified and H2BK120C nucleosome reconstitutions were done as described ( Armache et al., 2011 ; Dyer et al., 2004 ). Briefly, to assemble recombinant histone octamers, equimolar amounts of each of the 4 histones were mixed and dialyzed into refolding buffer. Octamers were purified by size exclusion chromatography on a Superdex 200 column (GE healthcare) in refolding buffer. Nucleosomes were assembled by mixing purified Widom 601 DNA and histone octamers and dialyzing overnight with gradient salt dialysis using a peristaltic pump (Gilson Rapid Pump). Assembled nucleosomes were purified through a Resource Q ion exchange column (GE Healthcare). Purified nucleosomes were dialyzed into TCS buffer (20 mM Tris-HCl pH 7.5, 1 mM EDTA, 1 mM DTT), concentrated, and stored at 4°C until use.

Gradient Fixation

(GraFix) of Dot1L with the nucleosome To prepare GraFix sample, assembled nucleosome was mixed with purified Dot1L protein in the molar ratio of 1:10 and dialyzed into Buffer A (Buffer A: 20 mM HEPES pH 7.9. 150 mM KCl, 5% Glycerol, 1 mM DTT) for 3 hours in 4°C, then supplemented with 5× molar ratio of SAH (Sigma) (H2BK120Ub nucleosome) and incubated on ice for 20 minutes before GraFix. Buffer B and buffer C (Buffer B: 20 mM HEPES pH 7.0, 150 mM KCl, 10% Glycerol, 1 mM DTT, 1 mM MgCl 2 / Buffer C: 20 mM HEPES pH 7.0, 150 mM KCl, 30% Glycerol, 1 mM DTT, 1 mM MgCl 2 , 0.1% Glutaraldehyde) were mixed using gradient maker (Gradient Master, Biocomp instrument). Prepared sample was then added on the top layer of the tube mixture, and centrifuged overnight (Beckman Coulter Optima XL-100K). Solution in the tube was then fractionated and analyzed. Selected fractions were dialyzed into dialysis buffer (Dialysis Buffer: 20 mM HEPES 7.0, 150 mM KCl, 1 mM MgCl 2 , 1 mM DTT) and concentrated.

Nucleosome binding assay

Increasing amounts of catalytic domains of Dot1L or yeast Dot1 proteins (50 nM to 1000 nM) were incubated with 12 nM recombinant nucleosomes in EMSA buffer (10 mM Tris pH 7.5, 100 mM NaCl, 2.5% Glycerol and 1 mM DTT) at room temperature for 30 min. The binding reactions were resolved on native polyacrylamide gels (6% PAGE, 0.2 × TBE), stained with SYBR™ Gold (Thermo Fisher), visualized on a Typhoon Trio+ scanner (Molecular Dynamics), and quantified using the program ImageQuant 5.2v (Molecular Dynamics). The amount of Dot1L or yeast Dot1 bound to nucleosomes was determined by measuring the decrease in free nucleosome in each reaction. The background was subtracted from Dot1-free samples. The free DNA was taken under in consideration for the calculation of free nucleosome. The apparent K D and the Hill coefficient for each binding curve were calculated by fitting the specific binding with Hill slope equation using the program Prism 7 (GraphPad). The final parameters were calculated from at least 3 independent experiments. (n≥3/data point). Data were plotted as mean ± s.d.

Endpoint methylation assay

Dot1L methyltransferase activity was determined by the monitoring the production S-adenosyl homocysteine (SAH) in presence of SAM and nucleosomes. Briefly, 50 nM of wild-type or mutant Dot1L proteins were combined with 1 μM unmodified or H2B-K120Ub nucleosome in methyltransferase buffer (20 mM Tris pH 8.0, 50 mM NaCl, 1 mM DTT, 0.1 mg/ml BSA, 20 μM SAM) in a total reaction volume of 12 μl. The methyltransferase reactions were performed at 30°C for 30 min and stopped with 3 μl of 0.5% TFA. The SAH production was determined with MTase-Glo methyltransferase kit (Promega) by transferring 8 ul of the stopped reaction to a white 384 well plate, developing and measuring luminescence using an EnSpire 2300 Multilabel plate reader (Perkin Elmer). The assays were performed with 3 replicates. Histone Methyl Transferase assay (HMT) The reactions were performed by mixing the indicated amounts of recombinant human Dot1L (1 pmol at the highest concentration) and nucleosomes (10 pmol) in the HMT assay buffer (50 mM Tris-HCl pH 8.5, 50 mM NaCl, 5 mM MgCl 2 and 1 mM DTT) with 10 μM SAM at 30°C for 1 hour or 3 hours, in 25 μL final volume. The reactions were stopped by the addition of 5.0 μL 5× SDS buffer. The products were resolved in a 15% SDS-PAGE gel and transferred in to 0.2-μm polyvinylidene difluoride PDVF membrane (Bio-Rad). The histone methylation level was determined by incubating the membranes with anti-H3K79Me3 (Abcam Ab2621, 1:1000) or anti-H3K79Me2 (Abcam Ab3594, 1:1000) for 12 hours at 4°C. The western blots were developed using ECL reagent (Thermo Fisher), and imaged in the ChemiDoc system (Bio-Rad). The loading control was determined with FastBlue staining, or blotting with anti-H4 antibody (Abcam Ab7311, 1:1000).

Cryo-EM data acquisition and processing

Cryo-EM grids of the Dot1L-H2BK120Ub (Sample 1) and Dot1L-Nuc (Sample 2) complexes were prepared following established protocol ( Li et al., 2013 ). For sample 1, 3.0 uL of the sample at 0.5 mg/mL concentration was applied to a 30s glow-discharged Quantifoil gold grids (400 mesh, 1.2 um hole size), blotted for 3s using Vitrobot Mark IV (FEI Company) at 4°C and 100% humidity. Sample 2, 3.0 uL of the sample at 0.5 mg/mL was applied to Quantifoil holey carbon grids (200 mesh, 1.2 um hole size) glow discharged for 10 seconds and blotted for 1.5s using Vitrobot Mark III (FEI Company) at 18°C and 100% humidity. The grids were subsequently plunge-frozen in liquid ethane cooled to liquid nitrogen temperatures. Sample 1 data was collected on two microscopes: on FEI Titan Krios 300kV (A, Figure S1 ) using EPU and on FEI Arctica 200kV (B, Figure S2 ) with SerialEM ( Mastronarde, 2005 ); these datasets were processed separately. All images were recorded using Gatan K2 Summit direct electron detector camera at a nominal magnification of (A) 130,000× and (B) 28,000× (calibrated physical pixel size of 1.035 Å/pixel and 1.45 Å/pixel, respectively). The total exposure time was 10 seconds, and each image was fractioned into 50 subframes, each frame with an exposure time of 0.2 s. Per frame dose was 1.1 e/Å 2 (A, B) leading to a total accumulated dose of 44 electrons per Å 2 on the specimen. All Sample 1A images were recorded with a defocus in the range from 1.7 to 3.2 um, Sample 1B from 1.8 to 3.4. Sample 2 data was collected on a TF30 Polara (FEI) operated at 300 kV using SerialEM ( Figure S7 ). All images were recorded using a Gatan K2 Summit direct electron detector camera at a nominal magnification of 31,000× (calibrated physical pixel size of 1.22 Å/pixel). The total exposure time was 10 seconds, and each image was fractioned into 50 subframes, each frame with an exposure time of 0.2 s. Per frame dose was 1.2 e/Å 2 , leading to a total accumulated dose of 60 electrons per Å 2 on the specimen. All images were recorded with a defocus in the range from 1.7 to 3 um. Movie stacks acquired in super resolution mode were corrected for global and local motions (in 5×5 patches) using UCSF MotionCor2 v1.2.1 (Zheng et al., 2017), resulting in dose-weighted and un-weighted sums, and binned 2× using Fourier binning. The images were manually screened to eliminate empty or icy images. We used the non-dose weighted images for CTF estimation using GCTF ( Zhang, 2016 ); all subsequent processing was done using dose-weighted ones. We first picked a small number of particles manually (~2500), classified them into 6 reference-free 2D classes and subsequently used them with Gautomatch ( http://www.mrc-lmb.cam.ac.uk/kzhang/ ) for automated particle picking. Particles were then extracted from dose-weighted images using Relion3 ( Zivanov et al., 2018 ) and subjected to reference-free 2D classification into 50 classes using Cryosparc ( Punjani et al., 2017 ) to exclude picking artifacts. We then used Cryosparc's "Ab initio" option to generate a 3D initial model from the data. This yielded a reconstruction with a clearly defined nucleosome density, which was then used as a template in 3D refinement. For data homogenization, we employed 3D classification using Cryosparc’s "Ab initio" and "Heterogeneous refinement". The data was then transferred to Relion3, where after 3D classification, 3D autorefinement and postprocessing we employed CTF refinement and Beam tilt estimation. This led to a small improvement in resolution for Sample 1B, but had no influence on other datasets. Reconstruction with best scoring particles led to the best resolved Sample 1B 4.6 Å map. Subsequent final classification of Sample 1B provided us with slightly lower quality map (5.2 Å) , but with the least distorted complete catalytic domain of Dot1L. This, as well as all the final reconstructions of all the samples were done in cisTEM ( Grant et al., 2018 ). The detailed schemes of cleanup, classifications and refinements are shown in Figure S8 . All the information on the collected data and processing summaries are shown on Table S1 . Following the gold-standard refinement, the final resolutions were established using Fourier Shell Correlation (FSC) equals at 0.143 cutoff ( Rosenthal and Henderson, 2003 ).

Model building and refinement

Our cryo-EM reconstructions of the Dot1L-H2BK120Ub reveal that parts of Dot1L are flexible. To model the complex, we used the final reconstruction from Titan Krios ( Figure 1 and Figure S1 ) filtered it to 3.5 Å to get the global architecture of the complex (overall fit); we also filtered the same data to 3.3 Å and applied a mild sharpening to get the best quality of the interfaces with the ubiquitin and the nucleosome. The different maps used to build each region are summarized in Table S2 . Available X-ray crystal structures were used for initial rigid body fit into our cryo-EM 3.5 Å reconstruction. For the nucleosome, we used PDB 3TU4 ( Armache et al., 2011 ), 1NW3 ( Min et al., 2003 ) for catalytic domain of Dot1L and 1UBQ ( Vijay-Kumar et al., 1987 ) for ubiquitin, which were first manually fit into the density and then locally optimized using UCSF Chimera's "Fit in map" function ( Pettersen et al., 2004 ). Due to noisy map in this region, we used PDB 3QOX to place the SAH and obtained the restraints using eLBOW program in PHENIX. We then used Coot ( Emsley and Cowtan, 2004 ) for local adjustments of secondary elements and side-chains into densities, and subsequently, the complete model was refined using PHENIX (phenix.real_space_refine) ( Adams et al., 2010 ) using secondary structure, ADPs, rotamer and Ramachandran restraints in 100 iterations. It was then visually inspected, and Ramachandran outliers and problematic regions were fixed manually in Coot (final refinement statistics are summarized in Table S3 ). In addition to the catalytic domain of Dot1L that we constructed (5-332), we also see extra density at the end of the structured C-terminal end of Dot1L that could be a part of Dot1L; however, this region is too distorted in the cryo-EM reconstruction to reliably assign and model. To validate our structure, we first subjected the atoms to 0.1 Å displacement, and then refined it in phenix.real_space_refine against one of the half-maps. This refined model was then converted to a 3D density map and compared against two half-maps and the summed map. We calculated FSC curves with half-map 1 (used for refinement, "work", shown in our figures with blue), half-map 2 (not participating in refinement, hence "free", pink) and the summed map (green). Due to the dynamic nature of Dot1L, we decided to perform two validations. First, we tested the whole model against the 4.6 Å reconstruction ( Figure S2 ): This map contains the most complete density of the Dot1L catalytic domain. We also compared a truncated model (157-164 (ß5), 180-188 (ß6), 214-332 (remaining C-terminus) representing the best resolved part of Dot1L in the 3.5 Å reconstruction ( Figure S1 ). We see only small differences between the "work" and "free" FSC curves, which indicates lack of overfitting. Figures of the model and cryo-EM densities were prepared using Chimera, Coot and PyMOL. We also used the reconstruction from Talos Arctica filtered at 4.6 Å where the complete catalytic domain of Dot1L can be seen, to evaluate some additional regions of the complex including the densities for the second site of Dot1L and the UIM/K-rich ( Figure 2 and Figure S2 ). Crosslinking for mass spectrometry analysis 100 μg of nucleosomes and 550 μg of Dot1L were mixed, dialyzed into 20 mM HEPES pH7.9. 150 mM KCl, 5% Glycerol, 1 mM DTT buffer and crosslinked in 10-20 μl reactions containing 0.1-3 mM bis(sulfosuccinimidyl)suberate (BS3, Thermo Scientific) added from 10-fold stock solutions prepared with dialysis buffer. Crosslinking reactions were incubated on ice for 2 hours and then quenched with Tris-HCl pH 7.5 added to 50 mM final concentration from 0.5 M stock. For buffer exchange, proteins were precipitated by mixing reactions with 4 volumes of acetone (cooled at −20 °C) followed by 1-hour incubation at −20 °C. Proteins precipitates were collected by 10-min centrifugation at 16000 g at room temperature. Pellets were rinsed with 80% acetone, dried on air and dissolved in 50 μL of buffer containing 50 mM ammonium bicarbonate, 10 mM DTT and 2% SDS. After 10-min incubation at 60 °C samples were cooled at room temperature, mixed with 5 μL 0.5 M iodoacetamide and incubated for 30 min at room temperature in dark followed by second precipitation with acetone. Pellets were dissolved in 20 μL of denaturing buffer consisting of 50 mM ammonium bicarbonate and 8 M urea. For digestion, 5 μL of samples were diluted with 50 μL 50 mM ammonium bicarbonate containing 20 ng/μl trypsin/Lys-C mixture (Promega) and incubated overnight at 30°C. Digestion reactions were stopped by mixing with 5 μL 20% trifluoroacetic acid and clarified by 10-min centrifugation at 16000g. Peptides were desalted using C18 spin tips (Thermo Scientific) according to manufacturer's protocol, dried under vacuum and dissolved in 10 μL 0.1% formic acid.

Mass spectrometry

Peptides were analyzed in the Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific) coupled to Dionex UltiMate 3000 (Thermo Scientific) liquid chromatography system. Peptides were resolved on 50-cm long EASY-Spray PepMap RSLC C18 column using 80 min linear gradient from 96% buffer A (0.1% formic acid in water) to 40% buffer B (0.1% formic acid in acetonitrile) followed by 98% buffer B over 5 min with a flow rate of 300 nL/min. Each full MS scan (orbitrap analyzer, resolution 60,000) was followed by data-dependent MS/MS scans (orbitrap, resolution 15,000) for 20 top most abundant peptides after HCD fragmentation. Quadrupole isolation window was set to 2 m/z, precursors with charge state 4 to 6 were selected for fragmentation with collision energy set to 35. Monoisotopic precursor selection was enabled, and a dynamic exclusion window was set to 30 sec. Data processing and analysis. Raw data was processed with pLink2 ( http://pfind.ict.ac.cn/software/pLink/ ) using database of concatenated sequences of proteins constituting nucleosome, Dot1 and common contaminants. For peptides identification up to 4 missed trypsin cleavages were allowed, constant modification was set to carbamidomethyl at cysteine residue, variable modification was oxidation at methionine residue, cross-linker was set to BS3, FDR level was set to 1%. Other parameters were left unchanged. The crosslinking mass spectrometry diagrams ( Figure 5D and S3 ) were prepared in the xiNET server ( http://crosslinkviewer.org/ ). Analysis of H3K79 methylation. Methylation reactions were set up mixing 80 μg of Dot1L with 15 μg of Ub-nucleosome or non-Ub nucleosome in final volume 25 μl and incubated in dialysis at 4 °C for 2 hours. Separately, reactions with added SAM (10uM final concentration) served as positive control. After incubation reactions were mixed with 65 μl SDS buffer (50 mM ammonium bicarbonate, 10 mM DTT and 2% SDS). For negative control, Dot1 and nucleosomes were diluted with SDS buffer before mixing together. Quenched reactions and control samples were incubated at 60°C for 10 min, then cooled at room temperature followed by 30-min incubation with 50 mM iodoacetamide in dark. Alkylated proteins were precipitated with 4 volumes of cold acetone at −20 °C for 1 hour. Acetone pellets were dissolved in 20 μl of denaturing buffer (50 mM ammonium bicarbonate, 8 M urea). For digestion, 5 μl from each sample were mixed with 50 μl 50 mM ammonium bicarbonate containing 20 ng/μl trypsin/Lys-C mixture (Promega) and incubated at 30 °C overnight. Digestion reactions were quenched by mixing with 5 μl 20% trifluoroacetic acid and clarified by 10-min centrifugation at 16000g. Peptides were desalted using C18 spin tips (Thermo Scientific) according to manufacturer's protocol, dried under vacuum and dissolved in 10 μl 0.1% formic acid. Peptides were analyzed in the Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific) coupled to Dionex UltiMate 3000 (Thermo Scientific) liquid chromatography system. Peptides were resolved on 50-cm long EASY-Spray PepMap RSLC C18 column using 90-min gradient from 96% buffer A (0.1% formic acid in water) to 40% buffer B (0.1% formic acid in acetonitrile) followed by 98% buffer B over 5 min with a flow rate of 300 nl/min. Data-dependent acquisition method was set up as described elsewhere ( Davis et al., 2017 ). Each sample was analyzed twice. Analysis of data was done using Proteome Discoverer 2.1.1.21 (Thermo Fisher). Protein sequence database contained sequences of Dot1L, histones, ubiquitin, E. coli proteins and common contaminants. Sequest HT search engine was run with default parameters except that three missed trypsin cleavages were allowed and variable modifications were set to include phosphorylation of serine, threonine and tyrosine, acetylation of lysine and protein N-terminus, di- and trimethylation of lysine and arginine. Cysteine carbamidomethylation was set as a static modification. MS1 quantitation of peptides was done using Precursor Ions Area Detection module within Proteome Discoverer ( Figure S3 ).

QUANTIFICATION AND STATISTICAL ANALYSIS

Protein quantification was done by using an A280 extinction coefficient of 60,850 M −1 cm −1 for Dot1L WT or mutants; 60,975 M −1 cm −1 for Dot1L 1-332, 48,875 M −1 cm −1 for yeast Dot1 FL and 44,405 M −1 cm −1 for yeast Dot1Δ on a Nanodrop spectrophotometer (Thermo-Fisher).

DATA AND SOFTWARE AVAILABILITY

Our cryo-EM density maps were deposited in the Electron Microscopy Data Bank (EMDB) as EMD-0652 (Dot1L-H2BK120Ub Sample 1A, 3.5 Å), EMD-0653 (Dot1L-H2BK120Ub Sample 1B, 4.6 Å), EMD-0654 (Dot1L-H2BK120Ub Sample 1B, 5.2 Å) and EMD-0655 (Dot1L-Nuc Sample 2, 4.9Å) ( Table S1 ). Associated with the primary depositions were corresponding unsharpened map and both half maps. Atomic coordinates used for data analysis in our manuscript has been deposited in the Protein Data Bank (PDB) with accession code 6O96 ( Table S1 ). Raw EMSA gels images and Western Blots images have been deposited in the Mendeley Data repository: doi: 10.17632/59wbz6gbcp.1 . All data is available from the corresponding author upon reasonable request.

METHOD DETAILS Expression and purification of Dot1L and mutants Plasmid DNA of pET28-MHL-Dot1L (1-420) was purchased from Addgene (#40736), and respective Dot1L mutants were generated in lab using Q5 mutagenesis kit (NEB). Expression and purification steps for Dot1L and its mutants were the same. Dot1L plasmid DNA was transformed into ONESHOT™BL21(DE3) (ThermoFisher) competent cells and grown in 2xYT-Kan media. Dot1L was expressed as soluble protein by induction with 0.5 mM IPTG for 3 hours at 37°C upon the culture reaching OD600 = 0.4-0.6. Cells were harvested (Sorvall LYNX6000) and lysed (AvestinEmulsiflexC3), and the protein was purified through Ni-NTA agarose beads (Qiagen) (Lysis Buffer: 500 mM NaCl, 50 mM Tris pH 8.0, 5 mM Imidazole, 5% Glycerol, 2mM β-Mercaptoethanol (BME), 1× Protease Inhibitor (cOmplete EDTA free, Roche) / Elution Buffer: 500 mM NaCl, 50 mM Tris pH 8.0, 300 mM Imidazole, 5% Glycerol, 2 mM BME). Eluted Dot1L protein was digested with TEV protease (NEB) while dialyzing in digestion buffer (Digestion Buffer: 75 mM NaCl, 20 mM Tris pH 8.0, 5% Glycerol, 2 mM BME) overnight at 40°C to cleave off Polyhistidine-tag. After tag was cleaved off, sample was purified through HiTrap SP HP (GE Healthcare) liquid chromatography column (Buffer A: 75 mM NaCl, 25 mM HEPES pH 7.5, 5% glycerol, 2 mM BME / Buffer B: 1 M NaCl, 25 mM HEPES pH 7.5, 5% glycerol, 2 mM BME). Selected fractions were further purified through HiLoad Superdex 200 16/600 (GE Healthcare) size-exclusion liquid chromatography column (S200 Buffer: 150 mM NaCl, 10 mM HEPES pH 7.5, 2 mM DTT). Purified Dot1L protein was then concentrated, flash frozen in liquid nitrogen and stored in −80°C for future use.

Expression and purification of yeast Dot1 Plasmid

DNA of pET15b-His-3C-Dot1 (158-582) was generated using Gibson Assembly (NEB).

Yeast Dot1 plasmid

DNA was transformed into ONESHOT™BL21(DE3) (ThermoFisher) competent cells and grown in 2xYT-Amp media. Yeast Dot1 was expressed as soluble protein by inducing with 0.1mM IPTG for 16 hours at 18°C upon the culture reaching OD600 = 0.4-0.6. Cells were harvested (Sorvall LYNX6000) and lysed (AvestinEmulsiflexC3), and the protein was purified through Ni-NTA agarose beads (Qiagen) (Lysis Buffer: 500 mM NaCl, 20 mM HEPES pH 7.5, 20 mM Imidazole, 1 mM BME, 1× Protease Inhibitor / Elution Buffer: 500 mM NaCl, 20 mM HEPES pH 7.5, 300 mM Imidazole, 1 mM BME). Eluted yeast Dot1 protein was digested with 3C PreScission Protease (GE Healthcare) while dialyzed in digestion buffer (Digestion Buffer: 100 mM NaCl, 20 mM HEPES pH 7.5, 1 mM DTT) overnight at 4°. After which, sample was purified through HiTrap SP HP (GE Healthcare) liquid chromatography column (Buffer A: 100 mM NaCl, 20 mM HEPES pH 7.5, 1 mM DTT / Buffer B: 1 M NaCl, 20 mM HEPES pH 7.5, 1 mM DTT). Selected fractions were further purified through HiLoad Superdex 200 (GE Healthcare) size-exclusion liquid chromatography column (S200 Buffer: 150 mM NaCl, 10 mM Tris pH 7.5, 5 mM DTT). Purified yeast Dot1 protein was then concentrated, flash frozen in liquid nitrogen and stored in −80°C for future use.

Expression and purification of Ubiquitin Plasmid DNA of pET-His-UB

G76C was a generous gift from Dr. Tingting Yao. Ubiquitin plasmid DNA was transformed into SoluBL21™ (amsbio) competent cells and grown in 2xYT-Amp Media. Ubiquitin was expressed as soluble protein by inducing with 0.4 mM IPTG for 4 hours at 37°C upon the culture reaching OD600 = 0.4-0.6. Bacteria cells were harvested and lysed (AvestinEmulsiflexC3). Protein was then purified through Ni-NTA agarose beads (Qiagen) (Lysis Buffer: 300 mM NaCl, 50 mM Tris pH 8.0, 10 mM Imidazole, 5 mM BME, 1× Protease Inhibitor / Elution Buffer: 300 mM NaCl, 50 mM Tris pH 8.0, 300 mM Imidazole, 5 mM BME) followed by HiTrap Q HP (GE Healthcare) liquid chromatography column (Buffer A: 50 mM NaCl, 20 mM Tris pH 8.0, 0.2 mM EDTA, 10 mM BME / Buffer B: 1 M NaCl, 20 mM Tris pH 8.0, 0.2 mM EDTA, 10 mM BME). Purified ubiquitin was then dialyzed against water supplemented with 1 mM acetic acid followed by flash freezing in liquid nitrogen and lyophilized using VIRTISSentry. Purification of Widom 601 DNA A plasmid containing 8 copies of the Widom 601 nucleosome positioning sequence, each flanked by the EcoRV restriction enzyme cutting site ( Armache et al., 2011 ), was transformed into DH5α™ (ThermoFisher) competent cells and grown in 2xYT-Amp media overnight. The 601 DNA fragment was excised using EcoRV and purified using previously published protocols ( Dyer et al., 2004 ).

Expression and purification of wild-type Xenopus

Histones and Histone mutant H2BK120C Plasmids containing wild-type Xenopus histones were a generous gift from Dr. Karolin Luger, and mutant histone H2B-K120C was generated using the Q5 mutagenesis kit (NEB). Briefly, each histone was expressed in Rosetta (DE3) cells (Novagen), extracted from inclusion bodies, and purified sequentially by size exclusion and anion chromatography using previously published protocols ( Dyer et al., 2004 ). Purified histones were freeze dried using a Sentry lyophilizer (VirTis). Ubiquitination of Histone H2BK120C We followed previously published protocol ( Long et al., 2014 ). Briefly, lyophilized ubiquitin and histone H2BK120C were re-suspended (Re-suspension buffer: 10 mM Acetic acid, 7M Urea-Deionized) and mixed in the ratio of 2:1. Sodium Tetraborate, Urea and TCEP were added to achieve final concentration of 50 mM, 6M and 5 mM, respectively. Final solution was incubated at room-temperature for 30 minutes. Then, an amount of crosslinker (Di-Chloracetone diluted in Di-Methylformamide) equal to one-half molar ratio of total sulfhydryl groups was added to the solution and incubated on ice for additional 30 minutes. Reaction was ended by addition of BME to final concentration of 5 mM. Solution was then diluted 10 times with Denaturing Binding Buffer (Denaturing Binding Buffer: 50 mM Sodium phosphate (NaPi), 50 mM Tris pH 8.0, 300 mM NaCl, 6 M Urea, 10 mM Imidazole, 5 mM BME) and purified through Ni-NTA agarose beads (Qiagen) (Denaturing Elution Buffer: 50 mM NaPi, 50 mM Tris pH 8.0, 300 mM NaCl, 6 M Urea, 250 mM Imidazole, 5 mM BME). Purified Ubiquitinated-H2BK120C was dialyzed into water supplemented with 1 mM BME and lyophilized using VIRTISSentry. Reconstitution of nucleosomes Unmodified and H2BK120C nucleosome reconstitutions were done as described ( Armache et al., 2011 ; Dyer et al., 2004 ). Briefly, to assemble recombinant histone octamers, equimolar amounts of each of the 4 histones were mixed and dialyzed into refolding buffer. Octamers were purified by size exclusion chromatography on a Superdex 200 column (GE healthcare) in refolding buffer. Nucleosomes were assembled by mixing purified Widom 601 DNA and histone octamers and dialyzing overnight with gradient salt dialysis using a peristaltic pump (Gilson Rapid Pump). Assembled nucleosomes were purified through a Resource Q ion exchange column (GE Healthcare). Purified nucleosomes were dialyzed into TCS buffer (20 mM Tris-HCl pH 7.5, 1 mM EDTA, 1 mM DTT), concentrated, and stored at 4°C until use.

Gradient Fixation

(GraFix) of Dot1L with the nucleosome To prepare GraFix sample, assembled nucleosome was mixed with purified Dot1L protein in the molar ratio of 1:10 and dialyzed into Buffer A (Buffer A: 20 mM HEPES pH 7.9. 150 mM KCl, 5% Glycerol, 1 mM DTT) for 3 hours in 4°C, then supplemented with 5× molar ratio of SAH (Sigma) (H2BK120Ub nucleosome) and incubated on ice for 20 minutes before GraFix. Buffer B and buffer C (Buffer B: 20 mM HEPES pH 7.0, 150 mM KCl, 10% Glycerol, 1 mM DTT, 1 mM MgCl 2 / Buffer C: 20 mM HEPES pH 7.0, 150 mM KCl, 30% Glycerol, 1 mM DTT, 1 mM MgCl 2 , 0.1% Glutaraldehyde) were mixed using gradient maker (Gradient Master, Biocomp instrument). Prepared sample was then added on the top layer of the tube mixture, and centrifuged overnight (Beckman Coulter Optima XL-100K). Solution in the tube was then fractionated and analyzed. Selected fractions were dialyzed into dialysis buffer (Dialysis Buffer: 20 mM HEPES 7.0, 150 mM KCl, 1 mM MgCl 2 , 1 mM DTT) and concentrated.

Nucleosome binding assay

Increasing amounts of catalytic domains of Dot1L or yeast Dot1 proteins (50 nM to 1000 nM) were incubated with 12 nM recombinant nucleosomes in EMSA buffer (10 mM Tris pH 7.5, 100 mM NaCl, 2.5% Glycerol and 1 mM DTT) at room temperature for 30 min. The binding reactions were resolved on native polyacrylamide gels (6% PAGE, 0.2 × TBE), stained with SYBR™ Gold (Thermo Fisher), visualized on a Typhoon Trio+ scanner (Molecular Dynamics), and quantified using the program ImageQuant 5.2v (Molecular Dynamics). The amount of Dot1L or yeast Dot1 bound to nucleosomes was determined by measuring the decrease in free nucleosome in each reaction. The background was subtracted from Dot1-free samples. The free DNA was taken under in consideration for the calculation of free nucleosome. The apparent K D and the Hill coefficient for each binding curve were calculated by fitting the specific binding with Hill slope equation using the program Prism 7 (GraphPad). The final parameters were calculated from at least 3 independent experiments. (n≥3/data point). Data were plotted as mean ± s.d.

Endpoint methylation assay

Dot1L methyltransferase activity was determined by the monitoring the production S-adenosyl homocysteine (SAH) in presence of SAM and nucleosomes. Briefly, 50 nM of wild-type or mutant Dot1L proteins were combined with 1 μM unmodified or H2B-K120Ub nucleosome in methyltransferase buffer (20 mM Tris pH 8.0, 50 mM NaCl, 1 mM DTT, 0.1 mg/ml BSA, 20 μM SAM) in a total reaction volume of 12 μl. The methyltransferase reactions were performed at 30°C for 30 min and stopped with 3 μl of 0.5% TFA. The SAH production was determined with MTase-Glo methyltransferase kit (Promega) by transferring 8 ul of the stopped reaction to a white 384 well plate, developing and measuring luminescence using an EnSpire 2300 Multilabel plate reader (Perkin Elmer). The assays were performed with 3 replicates. Histone Methyl Transferase assay (HMT) The reactions were performed by mixing the indicated amounts of recombinant human Dot1L (1 pmol at the highest concentration) and nucleosomes (10 pmol) in the HMT assay buffer (50 mM Tris-HCl pH 8.5, 50 mM NaCl, 5 mM MgCl 2 and 1 mM DTT) with 10 μM SAM at 30°C for 1 hour or 3 hours, in 25 μL final volume. The reactions were stopped by the addition of 5.0 μL 5× SDS buffer. The products were resolved in a 15% SDS-PAGE gel and transferred in to 0.2-μm polyvinylidene difluoride PDVF membrane (Bio-Rad). The histone methylation level was determined by incubating the membranes with anti-H3K79Me3 (Abcam Ab2621, 1:1000) or anti-H3K79Me2 (Abcam Ab3594, 1:1000) for 12 hours at 4°C. The western blots were developed using ECL reagent (Thermo Fisher), and imaged in the ChemiDoc system (Bio-Rad). The loading control was determined with FastBlue staining, or blotting with anti-H4 antibody (Abcam Ab7311, 1:1000).

Cryo-EM data acquisition and processing

Cryo-EM grids of the Dot1L-H2BK120Ub (Sample 1) and Dot1L-Nuc (Sample 2) complexes were prepared following established protocol ( Li et al., 2013 ). For sample 1, 3.0 uL of the sample at 0.5 mg/mL concentration was applied to a 30s glow-discharged Quantifoil gold grids (400 mesh, 1.2 um hole size), blotted for 3s using Vitrobot Mark IV (FEI Company) at 4°C and 100% humidity. Sample 2, 3.0 uL of the sample at 0.5 mg/mL was applied to Quantifoil holey carbon grids (200 mesh, 1.2 um hole size) glow discharged for 10 seconds and blotted for 1.5s using Vitrobot Mark III (FEI Company) at 18°C and 100% humidity. The grids were subsequently plunge-frozen in liquid ethane cooled to liquid nitrogen temperatures. Sample 1 data was collected on two microscopes: on FEI Titan Krios 300kV (A, Figure S1 ) using EPU and on FEI Arctica 200kV (B, Figure S2 ) with SerialEM ( Mastronarde, 2005 ); these datasets were processed separately. All images were recorded using Gatan K2 Summit direct electron detector camera at a nominal magnification of (A) 130,000× and (B) 28,000× (calibrated physical pixel size of 1.035 Å/pixel and 1.45 Å/pixel, respectively). The total exposure time was 10 seconds, and each image was fractioned into 50 subframes, each frame with an exposure time of 0.2 s. Per frame dose was 1.1 e/Å 2 (A, B) leading to a total accumulated dose of 44 electrons per Å 2 on the specimen. All Sample 1A images were recorded with a defocus in the range from 1.7 to 3.2 um, Sample 1B from 1.8 to 3.4. Sample 2 data was collected on a TF30 Polara (FEI) operated at 300 kV using SerialEM ( Figure S7 ). All images were recorded using a Gatan K2 Summit direct electron detector camera at a nominal magnification of 31,000× (calibrated physical pixel size of 1.22 Å/pixel). The total exposure time was 10 seconds, and each image was fractioned into 50 subframes, each frame with an exposure time of 0.2 s. Per frame dose was 1.2 e/Å 2 , leading to a total accumulated dose of 60 electrons per Å 2 on the specimen. All images were recorded with a defocus in the range from 1.7 to 3 um. Movie stacks acquired in super resolution mode were corrected for global and local motions (in 5×5 patches) using UCSF MotionCor2 v1.2.1 (Zheng et al., 2017), resulting in dose-weighted and un-weighted sums, and binned 2× using Fourier binning. The images were manually screened to eliminate empty or icy images. We used the non-dose weighted images for CTF estimation using GCTF ( Zhang, 2016 ); all subsequent processing was done using dose-weighted ones. We first picked a small number of particles manually (~2500), classified them into 6 reference-free 2D classes and subsequently used them with Gautomatch ( http://www.mrc-lmb.cam.ac.uk/kzhang/ ) for automated particle picking. Particles were then extracted from dose-weighted images using Relion3 ( Zivanov et al., 2018 ) and subjected to reference-free 2D classification into 50 classes using Cryosparc ( Punjani et al., 2017 ) to exclude picking artifacts. We then used Cryosparc's "Ab initio" option to generate a 3D initial model from the data. This yielded a reconstruction with a clearly defined nucleosome density, which was then used as a template in 3D refinement. For data homogenization, we employed 3D classification using Cryosparc’s "Ab initio" and "Heterogeneous refinement". The data was then transferred to Relion3, where after 3D classification, 3D autorefinement and postprocessing we employed CTF refinement and Beam tilt estimation. This led to a small improvement in resolution for Sample 1B, but had no influence on other datasets. Reconstruction with best scoring particles led to the best resolved Sample 1B 4.6 Å map. Subsequent final classification of Sample 1B provided us with slightly lower quality map (5.2 Å) , but with the least distorted complete catalytic domain of Dot1L. This, as well as all the final reconstructions of all the samples were done in cisTEM ( Grant et al., 2018 ). The detailed schemes of cleanup, classifications and refinements are shown in Figure S8 . All the information on the collected data and processing summaries are shown on Table S1 . Following the gold-standard refinement, the final resolutions were established using Fourier Shell Correlation (FSC) equals at 0.143 cutoff ( Rosenthal and Henderson, 2003 ).

Model building and refinement

Our cryo-EM reconstructions of the Dot1L-H2BK120Ub reveal that parts of Dot1L are flexible. To model the complex, we used the final reconstruction from Titan Krios ( Figure 1 and Figure S1 ) filtered it to 3.5 Å to get the global architecture of the complex (overall fit); we also filtered the same data to 3.3 Å and applied a mild sharpening to get the best quality of the interfaces with the ubiquitin and the nucleosome. The different maps used to build each region are summarized in Table S2 . Available X-ray crystal structures were used for initial rigid body fit into our cryo-EM 3.5 Å reconstruction. For the nucleosome, we used PDB 3TU4 ( Armache et al., 2011 ), 1NW3 ( Min et al., 2003 ) for catalytic domain of Dot1L and 1UBQ ( Vijay-Kumar et al., 1987 ) for ubiquitin, which were first manually fit into the density and then locally optimized using UCSF Chimera's "Fit in map" function ( Pettersen et al., 2004 ). Due to noisy map in this region, we used PDB 3QOX to place the SAH and obtained the restraints using eLBOW program in PHENIX. We then used Coot ( Emsley and Cowtan, 2004 ) for local adjustments of secondary elements and side-chains into densities, and subsequently, the complete model was refined using PHENIX (phenix.real_space_refine) ( Adams et al., 2010 ) using secondary structure, ADPs, rotamer and Ramachandran restraints in 100 iterations. It was then visually inspected, and Ramachandran outliers and problematic regions were fixed manually in Coot (final refinement statistics are summarized in Table S3 ). In addition to the catalytic domain of Dot1L that we constructed (5-332), we also see extra density at the end of the structured C-terminal end of Dot1L that could be a part of Dot1L; however, this region is too distorted in the cryo-EM reconstruction to reliably assign and model. To validate our structure, we first subjected the atoms to 0.1 Å displacement, and then refined it in phenix.real_space_refine against one of the half-maps. This refined model was then converted to a 3D density map and compared against two half-maps and the summed map. We calculated FSC curves with half-map 1 (used for refinement, "work", shown in our figures with blue), half-map 2 (not participating in refinement, hence "free", pink) and the summed map (green). Due to the dynamic nature of Dot1L, we decided to perform two validations. First, we tested the whole model against the 4.6 Å reconstruction ( Figure S2 ): This map contains the most complete density of the Dot1L catalytic domain. We also compared a truncated model (157-164 (ß5), 180-188 (ß6), 214-332 (remaining C-terminus) representing the best resolved part of Dot1L in the 3.5 Å reconstruction ( Figure S1 ). We see only small differences between the "work" and "free" FSC curves, which indicates lack of overfitting. Figures of the model and cryo-EM densities were prepared using Chimera, Coot and PyMOL. We also used the reconstruction from Talos Arctica filtered at 4.6 Å where the complete catalytic domain of Dot1L can be seen, to evaluate some additional regions of the complex including the densities for the second site of Dot1L and the UIM/K-rich ( Figure 2 and Figure S2 ). Crosslinking for mass spectrometry analysis 100 μg of nucleosomes and 550 μg of Dot1L were mixed, dialyzed into 20 mM HEPES pH7.9. 150 mM KCl, 5% Glycerol, 1 mM DTT buffer and crosslinked in 10-20 μl reactions containing 0.1-3 mM bis(sulfosuccinimidyl)suberate (BS3, Thermo Scientific) added from 10-fold stock solutions prepared with dialysis buffer. Crosslinking reactions were incubated on ice for 2 hours and then quenched with Tris-HCl pH 7.5 added to 50 mM final concentration from 0.5 M stock. For buffer exchange, proteins were precipitated by mixing reactions with 4 volumes of acetone (cooled at −20 °C) followed by 1-hour incubation at −20 °C. Proteins precipitates were collected by 10-min centrifugation at 16000 g at room temperature. Pellets were rinsed with 80% acetone, dried on air and dissolved in 50 μL of buffer containing 50 mM ammonium bicarbonate, 10 mM DTT and 2% SDS. After 10-min incubation at 60 °C samples were cooled at room temperature, mixed with 5 μL 0.5 M iodoacetamide and incubated for 30 min at room temperature in dark followed by second precipitation with acetone. Pellets were dissolved in 20 μL of denaturing buffer consisting of 50 mM ammonium bicarbonate and 8 M urea. For digestion, 5 μL of samples were diluted with 50 μL 50 mM ammonium bicarbonate containing 20 ng/μl trypsin/Lys-C mixture (Promega) and incubated overnight at 30°C. Digestion reactions were stopped by mixing with 5 μL 20% trifluoroacetic acid and clarified by 10-min centrifugation at 16000g. Peptides were desalted using C18 spin tips (Thermo Scientific) according to manufacturer's protocol, dried under vacuum and dissolved in 10 μL 0.1% formic acid.

Mass spectrometry

Peptides were analyzed in the Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific) coupled to Dionex UltiMate 3000 (Thermo Scientific) liquid chromatography system. Peptides were resolved on 50-cm long EASY-Spray PepMap RSLC C18 column using 80 min linear gradient from 96% buffer A (0.1% formic acid in water) to 40% buffer B (0.1% formic acid in acetonitrile) followed by 98% buffer B over 5 min with a flow rate of 300 nL/min. Each full MS scan (orbitrap analyzer, resolution 60,000) was followed by data-dependent MS/MS scans (orbitrap, resolution 15,000) for 20 top most abundant peptides after HCD fragmentation. Quadrupole isolation window was set to 2 m/z, precursors with charge state 4 to 6 were selected for fragmentation with collision energy set to 35. Monoisotopic precursor selection was enabled, and a dynamic exclusion window was set to 30 sec. Data processing and analysis. Raw data was processed with pLink2 ( http://pfind.ict.ac.cn/software/pLink/ ) using database of concatenated sequences of proteins constituting nucleosome, Dot1 and common contaminants. For peptides identification up to 4 missed trypsin cleavages were allowed, constant modification was set to carbamidomethyl at cysteine residue, variable modification was oxidation at methionine residue, cross-linker was set to BS3, FDR level was set to 1%. Other parameters were left unchanged. The crosslinking mass spectrometry diagrams ( Figure 5D and S3 ) were prepared in the xiNET server ( http://crosslinkviewer.org/ ). Analysis of H3K79 methylation. Methylation reactions were set up mixing 80 μg of Dot1L with 15 μg of Ub-nucleosome or non-Ub nucleosome in final volume 25 μl and incubated in dialysis at 4 °C for 2 hours. Separately, reactions with added SAM (10uM final concentration) served as positive control. After incubation reactions were mixed with 65 μl SDS buffer (50 mM ammonium bicarbonate, 10 mM DTT and 2% SDS). For negative control, Dot1 and nucleosomes were diluted with SDS buffer before mixing together. Quenched reactions and control samples were incubated at 60°C for 10 min, then cooled at room temperature followed by 30-min incubation with 50 mM iodoacetamide in dark. Alkylated proteins were precipitated with 4 volumes of cold acetone at −20 °C for 1 hour. Acetone pellets were dissolved in 20 μl of denaturing buffer (50 mM ammonium bicarbonate, 8 M urea). For digestion, 5 μl from each sample were mixed with 50 μl 50 mM ammonium bicarbonate containing 20 ng/μl trypsin/Lys-C mixture (Promega) and incubated at 30 °C overnight. Digestion reactions were quenched by mixing with 5 μl 20% trifluoroacetic acid and clarified by 10-min centrifugation at 16000g. Peptides were desalted using C18 spin tips (Thermo Scientific) according to manufacturer's protocol, dried under vacuum and dissolved in 10 μl 0.1% formic acid. Peptides were analyzed in the Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific) coupled to Dionex UltiMate 3000 (Thermo Scientific) liquid chromatography system. Peptides were resolved on 50-cm long EASY-Spray PepMap RSLC C18 column using 90-min gradient from 96% buffer A (0.1% formic acid in water) to 40% buffer B (0.1% formic acid in acetonitrile) followed by 98% buffer B over 5 min with a flow rate of 300 nl/min. Data-dependent acquisition method was set up as described elsewhere ( Davis et al., 2017 ). Each sample was analyzed twice. Analysis of data was done using Proteome Discoverer 2.1.1.21 (Thermo Fisher). Protein sequence database contained sequences of Dot1L, histones, ubiquitin, E. coli proteins and common contaminants. Sequest HT search engine was run with default parameters except that three missed trypsin cleavages were allowed and variable modifications were set to include phosphorylation of serine, threonine and tyrosine, acetylation of lysine and protein N-terminus, di- and trimethylation of lysine and arginine. Cysteine carbamidomethylation was set as a static modification. MS1 quantitation of peptides was done using Precursor Ions Area Detection module within Proteome Discoverer ( Figure S3 ).

Supplementary Material 2