Article pubs.acs.org/ac
Precision Mapping of Coexisting Modifications in Histone H3 Tails from Embryonic Stem Cells by ETD-MS/MS Hye Ryung Jung,† Simone Sidoli,† Simon Haldbo,† Richard R. Sprenger,† Veit Schwam ̈ mle,† ‡ ‡,§ ,† Diego Pasini, Kristian Helin, and Ole N. Jensen* †
Centre for Epigenetics, Department of Biochemistry and Molecular Biology, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark ‡ Centre for Epigenetics, Biotech Research and Innovation Centre, University of Copenhagen, Ole Maaløes Vej 5, DK-2200 Copenhagen, Denmark § The Danish Stem Cell Center, University of Copenhagen, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark S Supporting Information *
ABSTRACT: Post-translational modifications (PTMs) of histones play a major role in regulating chromatin dynamics and influence processes such as transcription and DNA replication. Here, we report 114 distinct combinations of coexisting PTMs of histone H3 obtained from mouse embryonic stem (ES) cells. Histone H3 N-terminal tail peptides (amino acids 1−50, 5−6 kDa) were separated by optimized weak cation exchange/hydrophilic interaction liquid chromatography (WCX/HILIC) and sequenced online by electron transfer dissociation (ETD) tandem mass spectrometry (MS/MS). High mass accuracy and near complete sequence coverage allowed unambiguous mapping of the major histone marks and discrimination between isobaric and nearly isobaric PTMs such as trimethylation and acetylation. Hierarchical data analysis identified H3K27me2-H3K36me2 as the most frequently observed PTMs in H3. Modifications at H3 residues K27 and K36 often coexist with the abundant mark K23ac, and we identified two frequently occurring quadruplet marks ‘K9me1K23acK27me2K36me2’ and ‘K9me3K23acK27me2K36me’, which might indicate a role in crosstalk. Co-occurrence frequency analysis revealed also an interplay between methylations of K9, K27, and K36, suggesting interdependence between histone methylation marks. We hypothesize that the most abundant coexisting PTMs may provide a signature for the permissive state of mouse ES cells.
C
differentiation, the bivalent domain is resolved to maintain either the repressive mark or the activation mark, thereby inducing or repressing transcription. Histone proteins assemble to form the nucleosome by the deposition of two H3-H4 dimers as a tetramer onto DNA followed by the deposition of two H2A-H2B dimers.7 Each of the four core histones share the common histone folding domain that is bound to DNA, while their basic N-terminal tails protrude from the nucleosomal core. These N-terminal histone tails can be heavily modified with phosphorylation (ph), monomethylation (me1), dimethylation (me2), trimethylation (me3), acetylation (ac), formylation, citrullination, and ubiquitylation (ubi) on lysine, arginine, serine, and threonine residues.2,8,9 These modifications create numerous marks of combinatorial PTMs that can be recognized by chromatin binding (non-histone) proteins,2,10 which is a basic concept of the ‘histone code’.11
oordinated modification of multiple, distinct amino acid residues in histone proteins is directly or indirectly involved in the activation or repression of genes1 and has emerged as an important mechanism for stabilizing a dynamic regulation of cellular processes and signaling to chromatin.1−3 Specific multisite modifications can act in a sequential and/or a combinatorial manner. Furthermore, crosstalk between posttranslational modifications (PTMs) may occur within a single protein (in cis) or between two or more proteins (in trans), as has been shown for p53, myocyte-specific enhancer factor 2A, and histones.1,2,4 Histones play a key role in diverse epigenetic phenomena such as differentiation and pluripotency, which is the developmental capacity of embryonic stem (ES) cells to differentiate to committed cells.5 The pluripotency state of ES cells is a result of dynamic and elaborate balance between combinations of histone modifications, mobilized transcription factors, and DNA methylation. For instance, chromatin immunoprecipitation studies revealed that key regulatory genes in ES cells are modified with a so-called ‘bivalent domain’ consisting of both a transcriptional activation mark, H3K4me3, and a transcriptional repressive mark, H3K27me3.6 Upon © XXXX American Chemical Society
Received: May 3, 2013 Accepted: July 28, 2013
A
dx.doi.org/10.1021/ac401299w | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
■
Alternatively, histone PTMs can alter intra- and internucleosomal interaction through changes of steric or charge interactions. For instance, reduction of a positive charge on H4K16 upon acetylation leads to decompression of chromatin.12 Traditionally histone PTMs were analyzed by antibodies for detecting a single PTM. Antibody-based technologies provided a better understanding of the localization of distinct histone PTMs on chromatin and the function of these modifications. However, antibody-based analysis suffers from epitope occlusion and crossreactivity, and few antibodies are available for detection of coexisting PTMs in histones. In contrast, mass spectrometry can sequence peptides and map multiple, coexisting modifications in proteins, including N-terminal histone tails and intact histones, and MS is now a workhorse technology for PTM characterization,13 although some limitations are still evident. For instance, mass spectrometers with low mass accuracy cannot accurately assign near isobaric modifications such as acetylation (42.011 Da) and trimethylation (42.047 Da) that are frequently observed on histone proteins. Therefore, it is extremely important to use high mass accuracy mass spectrometers for precise mapping of multisite histone PTMs. A majority of MS studies on histones have been performed using endoproteinases such as trypsin or ArgC to generate relatively small peptides (300−2000 Da), which are then analyzed by tandem mass spectrometry (MS/MS). It is a robust and sensitive method to map one or at most a few PTMs within a single peptide, but restricted in mapping combinatorial PTMs over longer stretches of the polypeptide chain. More recently, mass spectrometry was used to map combinatorial PTMs on intact proteins or large fragments of proteins, thereby detecting ‘long-distance’ combinations of modifications.14−17 Collision induced dissociation (CID) MS/MS is widely used for fragmentation of peptides. However, CID is not efficient enough to reliably obtain high sequence coverage by fragmentation of large (>3 kDa) peptide ions. More recently, higher energy collision induced dissociation (HCD) has gained popularity for peptide sequencing and PTM mapping, particularly for tryptic peptides.18 Electron-based fragmentation methods such as electron capture dissociation (ECD)19 and electron transfer dissociation (ETD)20 are wellsuited for efficient sequencing of large, multiply protonated peptides due to their fast and random fragmentation mechanism. Thus, these electron-based fragmentation methods combined with high mass accuracy tandem mass spectrometers were applied for mapping combinations of modifications on intact or long N-terminal tails of histones.14,15 The heterogeneous nature of histone proteins calls for efficient separation prior to sitespecific mapping of the high number of distinct combinations of modifications. Reversed phase (RP) liquid chromatography (LC) can separate histone subtypes to some extent,9,15 whereas more efficient separation of histone peptides is achieved by hydrophilic interaction liquid chromatography (HILIC) prior to the MS/MS analysis.16,17 Previously, HILIC combined with tandem mass spectrometry allowed mapping of more than 200 different combinations of modifications on H3 N-terminal tails derived from HeLa cells.21 Here, we adapted and optimized WCX/HILIC to separate 5−6 kDa H3 N-terminal tail peptides and interfaced it to high mass accuracy ETD tandem mass spectrometry (LTQ-Orbitrap XL/ETD) and a detailed data analysis workflow to reveal distinct combinatorial PTMs on histone H3 tails isolated from mouse ES cells.
Article
MATERIALS AND METHODS
Sample Preparation. Cellular nuclei were extracted from one 15-cm Petri dish of mouse ES cells as described previously.22 The nucleus solution was centrifuged at 10,000g for 10 min at 4 °C, and the pellet was resuspended in 400 μL of 0.2 M H2SO4 followed by thorough vortexing and incubation for 35 min at 4 °C. Samples were centrifuged at 16,000g for 10 min at 4 °C. Trichloroacetic acid (final 25%, Sigma) was added to the supernatant and incubated for 30 min at 4 °C, followed by centrifugation at 16,000g for 10 min at 4 °C. The pellet was washed once using ice-cold acetone (Merck) with 1% hydrochloric acid (Sigma), followed by two washes with ice-cold acetone. The pellet was air-dried and redissolved in 200 μL of ddH2O assisted by 5 min of sonication. Preparation of H3 N-Terminal Tail Peptides. A 200-μg portion of intact histones was separated by reversed phase (RP)HPLC using a C18 column (250 mm × 4.6 mm, Jupiter, 300 Å, Phenomenex, Torrance, CA, USA) on an Akta-Basic system (GE Healthcare). Solvent A and solvent B consisted of 0.06% TFA in ddH2O and 0.04% TFA and 90% MeCN (Sigma), respectively. The different histones were separated by increasing solvent B from 5% to 35% in 10 min, 35% to 55% in 80 min, and 60% to 90% in 2 min. The intact masses of the different histones were measured by MALDI TOF Bruker Ultraflex (Bruker Daltonics, Bremen, Germany) in the positive ion mode. One microliter of each histone-containing HPLC fraction was deposited with 0.5 μL of α-cyano-4-hydroxycinnamic acid (10 mg/mL, in 70% MeCN, 0.1% TFA) on a MALDI target plate. Then 40 μg of histone H3 was digested in 100 mM NH4HCO3 using endoproteinase GluC (Calbiochem, 1:20 enzyme to protein ratio) and incubated at room temperature for 4 h, followed by the addition of 1 μL of 1% TFA to quench the proteolytic activity. Digestion products were separated using a C18 RP-HPLC column (250 mm × 2 mm, Jupiter, 300 Å, Phenomenex, Torrance, CA, USA; 5% to 35% in 10 min, 35% to 50% in 60 min, and 50% to 90% in 2 min. Fractions of H3 tails were identified using MALDI TOF MS, and the corresponding fractions were freeze-dried and stored at −80 °C until further analysis. Weak Cation Exchange/Hydrophilic Interaction Liquid Chromatography. H3 long N-terminal tail peptides were analyzed by a WCX/HILIC-MS/MS method adapted from a previous study21 and further optimized by us. A WCX/HILIC column was packed with PolyCAT A, 1500 Å (TM 3 μm, cat. no. BMCT0315, PolyLC Inc). In addition to the analytical column (18 cm, 100 μm i.d., 360 μm o.d.), a trap column (1 cm, 100 μm i.d., 360 μm o.d.) was packed to establish two-column setup using an UltiMate 3000 (Dionex). Solvent A and solvent B were composed of 75% MeCN (Fisher Scientific), 20 mM propionic acid (Sigma), pH 6 and 25% MeCN, pH 2.5, respectively. pH values of solvent A and B were adjusted with ammonium hydroxide (Sigma) and formic acid (Merck) respectively. H3 Nterminal peptides were separated using a gradient of 60% to 70% or 55% to 65% in 55 min at 300 nL/min. Sample injection was 1 μg in 5 μL. Eluted peptides were directly injected into a LTQOrbitrap XL (Thermo Fisher Scientific, Bremen, Germany). Tandem Mass Spectrometry. The MS/MS spectra were acquired in a data-dependent mode automatically switching between MS and ETD-MS/MS. The nanoelectrospray ion source (Thermo Fisher Scientific) was used with a spray voltage of 1.5−2.0 kV. The capillary temperature was 200 °C, and no sweep, aux, and sheath gas was used. Each sample was analyzed twice by LC−MS/MS with different MS range: m/z 500−750 B
dx.doi.org/10.1021/ac401299w | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Figure 1. MS-based strategy for precision mapping of histone marks and distinct combinatorial PTMs on histone H3. (A) Schematic workflow representing the steps of sample preparation, analysis, and data processing. (B) A mass spectral heat map of H3 tails. Long H3 N-terminal peptides were separated by the number of acetyl-groups and by the number of methyl-groups. (C) Enlargement of the heatmap in the region of 1 acetylation.
serine and threonine (79.966 Da). Result files were then imported into an in-house developed software to verify correct assignment of PTMs. A spectral counting approach23 was applied to estimate the frequency and relative abundance of coexisting histone marks (PTMs); all identified spectra that passed the filters for correct assignment of PTMs were summed when they corresponded to the same peptide variant, considering as peptide variant a peptide with a unique combination of PTMs. The set of modified peptides identified in this manner was further analyzed by in-house developed software for data visualization and statistical assessment.
(+9 and +8 precursors) or m/z 750−1000 (+7 and +6 precursors). MS were obtained in the orbitrap with maximum ion injection time of 500 ms, auto gain control target 7 × 105, resolution 60000 and one microscan per MS. The two most abundant precursor ions within the inclusion list were selected for ETD with 80 ms reaction time. MS/MS spectra were measured by the orbitrap with maximum ion injection time 1000 ms, auto gain control target 1 × 105, resolution 30000 and 8 microscans per each MS/MS. As a consequence, the overall duty cycle was between 10 and 16 s. Charge states 1, 2, 3 were rejected from the precursor selection. Precursor isolation width was 2, and exclusion duration was 90 s. Data Analysis. The raw data from the LTQ-Orbitrap XL mass spectrometer were imported into the Proteome Discoverer 1.3 software (Thermo Fisher Scientific). Xtract (Thermo Fischer Scientific) was used as a process algorithm with a precursor selection minimum signal-to-noise 0. Processed data was searched using the mouse histone database exported from Uniprot (updated at 5/9/2011). The search was performed using Mascot v2.3 (Matrix Science, London, UK) as search engine with the following parameters: intact mass tolerance 2.2 Da, fragment mass tolerance: 0.01 Da; variable modifications: acetylation (42.011 Da) on N-terminal of proteins and lysine, mono-, dimethylation on lysine and arginine (14.016 and 28.031 Da), trimethylation on lysine (42.047 Da), phosphorylation on
■
RESULTS
Histones were recovered from mouse embryonic stem (ES) cells and separated by RP-HPLC (Figure 1A). A fraction containing histone H3 was digested with endoproteinase GluC which cleaves at the amide bond C-terminal to acid amino acids. In consequence, long N-terminal peptides covering amino acids (a.a.) 1−50 (5−6 kDa) were generated that can contain none, single or multiple site-specific modifications. The Glu-C digested histone samples were further separated by RP-HPLC. Next, fractions containing N-terminal histone H3 tails were analyzed by nanoliter flow WCX/HILIC interfaced to an LTQ-Orbitrap XL hybrid tandem mass spectrometer to characterize distinct combinations of site-specific protein modifications. Previously, C
dx.doi.org/10.1021/ac401299w | Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
Article
Figure 2. Hierarchy of PTM deposition on N-terminal tails of histone H3. Methylations were mainly located on amino acids near to the C-terminus in lightly modified peptides. (A) Single methylated peptides were observed to be heavily methylated on K27 and K36 (top), and a similar trend was observed for doubly methylated peptides (bottom). (B) Hierarchy in methylation deposition, estimated with the relative abundances of the combinatorial histone marks. H3 tails with only one PTM are mostly occupied the C-terminus, alternating mostly K27 (60%) and K36 (40%). As the number of modifications increased, amino acids near to the N-terminus were more frequently methylated.
WCX/HILIC was used to separate core and linker histone subtypes, thus allowing identification of individual histone isoforms.16,24 We adopted a WCX/HILIC method from a previous study,21 and further modified and optimized it to achieve robust and fast separation of polypeptides prior to the MS analysis. Here, we introduce a two-column set up combining a reversed-phase precolumn (trap) and an analytical HILIC column, instead of the previously reported one-column HILIC method.21 This new configuration allowed for easy and automated sample loading, instead of the more cumbersome and manual sample loading by using a pressure bomb. The run time was also reduced; the different types of H3 N-terminal peptides were separated within a relatively short time (