Article pubs.acs.org/ac
Fragment Ion Patchwork Quantification for Measuring Site-Specific Acetylation Degrees Rasha ElBashir,†,‡ Jens T. Vanselow,†,‡ Amelie Kraus,§ Christian J. Janzen,∥ T. Nicolai Siegel,§ and Andreas Schlosser*,‡ ‡
Rudolf Virchow Center for Experimental Biomedicine and §Research Center for Infectious Diseases, University of Wuerzburg, Josef-Schneider-Str. 2, 97080 Wuerzburg, Germany ∥ Department of Cell & Developmental Biology, Biocenter University of Wuerzburg, Am Hubland, 97074 Wuerzburg, Germany S Supporting Information *
ABSTRACT: We introduce fragment ion patchwork quantification as a new mass spectrometry-based approach for the highly accurate quantification of site-specific acetylation degrees. This method combines 13 C1-acetyl derivatization on the protein level, proteolysis by lowspecificity proteases and quantification on the fragment ion level. Acetylation degrees are determined from the isotope patterns of acetylated b and y ions. We show that this approach allows to determine site-specific acetylation degrees of all lysine residues for all core histones of Trypanosoma brucei. In addition, we demonstrate how this approach can be used to identify substrate sites of histone acetyltransferases.
L
D3-acetyl derivatization has been used in different applications for measuring acetylation degrees of histones and other proteins.7−12 More recently, D3-acetyl derivatization has been applied in combination with a targeted MS-based approach to study specific alterations of acetylation signatures in histone tails after depletion of HATs and HDACs.13 However, a significant limitation of D3-acetyl derivatization is that it causes retention time shifts between isotopologue peptides in C18 reversed-phase chromatography and thereby impairs quantification on the fragment ion level. Nevertheless, quantification on the fragment ion level is mandatory for the site-specific analysis of acetylation degrees within clusters of adjacent lysine residues, as typically found in the tail regions of histones. To overcome these limitations, we introduce “fragment ion patchwork quantification”, a new MS-based approach that allows determining site-specific acetylation degrees from fragment ion spectra with high accuracy, even within clusters of lysine residues.
ysine acetylation is an important post-translational modification (PTM) that is involved in the regulation of many cellular processes, such as DNA repair, cell cycle regulation, RNA splicing, RNA transcription, and cytoskeleton reorganization.1,2 Reversible acetylation of histone tails is a highly dynamic process maintained by histone acetyltransferases (HATs) and deacetylases (HDACs) and serves as an important regulator of gene expression.3 For the quantification of lysine acetylation, numerous mass spectrometry (MS)-based methods have been developed. Standard methods for relative quantification, such as SILAC4 or label-free quantification,5,6 have been applied for the quantitative analysis of acetylation sites. However, these methods can only be used for relative quantification, that is, measuring acetylated peptide foldchanges between samples, but cannot accurately determine site-specific acetylation degrees. The acetylation degree of a specific site is the fraction of this site being acetylated. If other types of modifications (e.g., methylation) at the same site are absent, the site-specific acetylation degree can be determined by measuring the molar ratio of the corresponding peptides in their acetylated and nonacetylated form. This however is a challenging task, because (i) acetylated and nonacetylated peptides are expected to have significant different ionization efficiencies and (ii) protease cleavage is influenced by acetylation at or close to the protease cleavage site. Both adversities can be overcome by applying chemical acetyl derivatization with stable isotope-labeled acetic anhydride on the protein level. This procedure eliminates the influence of acetylated sites on proteolytic cleavage and on ionization efficiencies of the generated peptides by converting all unmodified amino groups to acetylamino groups. Previously, © XXXX American Chemical Society
■
EXPERIMENTAL SECTION Chemicals and Proteases. Acetic anhydride-1,1′-13C2 (99 atom % 13C) was purchased from Cambridge Isotope Laboratories, sodium acetate-1-13C (99 atom % 13C) was ordered from Sigma-Aldrich. Sodium acetate (ACS reagent, ≥ 99.0%), acetic anhydride, ammonium bicarbonate, LC-MS grade acetonitrile (LC-MS CHROMASOLV), iodoacetamide (Bio ultra), formic acid (for MS), and L-cysteine were ordered Received: July 5, 2015 Accepted: September 3, 2015
A
DOI: 10.1021/acs.analchem.5b02517 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
column (25 cm × 75 μm ID, PepMap C-18 2 μm particles, 100 Å pore size) with either a 45 min (for recombinant histone H3) or a 120 min (for histone fractions obtained by acid extraction) linear gradient from 3 to 30% acetonitrile and 0.1% formic acid. Both MS and MS/MS scans were acquired in the Orbitrap analyzer with a resolution of 60.000. HCD fragmentation with 35% normalized collision energy was applied. A top speed datadependent MS/MS method with a fixed cycle time of 3 s was used. Dynamic exclusion was applied with a repeat count of 3 and an exclusion duration of 45 s. Minimum signal threshold for precursor selection was set to 50.000, and the width of the precursor selection window was set to 6 Da. Predictive AGC was used with a target value of 2 × 105 for MS scans and 5 × 104 MS/MS scans. EASY-IC was used for internal calibration. MS Raw Data Processing. Each raw data file was processed with Mascot Distiller 2.4 (Matrix Science) using two different types of parameter settings. One parameter set (ID) was used for generating peak lists for database searching with Mascot, the other parameter set (Quant) was used for generating peak lists for quantitative analysis (Table S-1). Database Searching. ID peak lists were searched with Mascot Server 2.4 against a custom database compiled from TriTrypDB.14 In addition to all histone sequences, the custom database contained the top 300 of proteins identified in our histone samples. Since acetylation with 13C1-acetyl is a proxy for in vivo unmodified lysines, we replaced all K for O in all protein sequences of the database fasta file. O was defined as a new amino acid with the chemical composition 13CC7H14N2O2 (monoisotopic mass 171.108883 Da). In order to take Nterminal acetylation into account we defined J as a new amino acid with the chemical composition of 13C1-acetyl (13CCH2O, monoisotopic mass 43.01392 Da). J was added to the Nterminus of all protein sequences of the database fasta file. The fasta file contained each protein sequence in two versions, one starting with J and the other starting with JM. All variable protein N-terminal and lysine modifications were defined relative to 13C1-acetyl (see below). In addition to carbamidomethyl at C (C2H3NO) as fixed modification, database searching was performed with the following variable modifications: oxidation at M (O), light acetyl at O (13C−1C), light acetyl at protein N-terminus (13C−1C), monomethyl at O (CH2), monomethyl at protein Nterminus (CH2), dimethyl at O (13C−1CH2O−1), dimethyl at protein N-terminus ( 13 C −1 CH 2 O −1 ), trimethyl at O ( 13C −1 C 2H 4 O−1 ), and trimethyl at protein N-terminus (13C−1C2H4O−1). These definitions take into account that monomethylated amino groups react with acetic anhydride, whereas di- and trimethylated amino groups do not react. Database searches were performed without protease specificity (enzyme: none), 7 ppm mass tolerance for the precursor, and 0.02 Da mass tolerance for fragment ions. Mascot search results were exported as XML file. In addition to standard settings, start and end positions of the peptides were included in the export file, peptides with score 0.99, maximum difference in the relative abundance of a single isotope peak
