Anal. Chem. 2006, 78, 4271-4280
Accelerated Articles
Quantitative Analysis of Modified Proteins and Their Positional Isomers by Tandem Mass Spectrometry: Human Histone H4 James J. Pesavento,† Craig A. Mizzen,‡,§ and Neil L. Kelleher,*,⊥
Center for Biophysics and Computational Biology, Department of Cell and Developmental Biology, Institute for Genomic Biology (IGB), Department of Chemistry, and The Center for Top Down Proteomics, University of Illinois at UrbanasChampaign, Urbana, Illinois 61801
Here we show that fragment ion abundances from dissociation of ions created from mixtures of multiply modified histone H4 (11 kDa) or of N-terminal synthetic peptides (2 kDa) correspond to their respective intact ion abundances measured by Fourier transform mass spectrometry. Isomeric mixtures of modified forms of the same protein are resolved and quantitated with a precision of e5% using the relative ratios of their fragment ions, with intact protein ions created by electrospray greatly easing many of the systematic biases that more strongly affect small peptides (e.g., differences in ionization efficiency and ion m/z values). The ion fragmentation methods validated here are directly extensible to intact human proteins to derive quantitative information on the highly related and often isomeric protein forms created by combinatorial arrays of posttranslational modifications.
Since its earliest days, the discipline of analytical chemistry has devised measurement tools that provide linear responses to the concentration of molecules in isolation or in mixtures. For mixtures of principal components in absorption spectroscopy, there is an “assumption of additivity” that largely holds true until the analyte signal from one molecule is affected by another or the S/N levels become too low to conform to Beer’s law.1 The classic example in mass spectrometry (MS) from the 1950s is the analysis of light hydrocarbon gases for quantitating both n-butane and isobutane and nearly 20 other components from the EI mass spectrum of their mixture because their individual spectra are linearly additive. As new ionization methods have allowed mixtures of higher mass components to be interrogated by high-resolution tandem mass spectrometry,2 a reinvestigation of the assumption * Corresponding author. Tel: (217) 244-3927. E-mail:
[email protected]. † Center for Biophysics and Computational Biology. ‡ Department of Cell and Developmental Biology. § Institute for Genomic Biology. | Department of Chemistry. ⊥ The Center for Top Down Proteomics. (1) Skoog, D. A.; Holler, F. J.; Nieman, T. A. Principles of Instrumental Analysis, 5th ed.; Saunders College Publishing: Orlando, FL, 1998. (2) McLafferty, F. W. Acc. Chem. Res. 1994, 27, 379-386. 10.1021/ac0600050 CCC: $33.50 Published on Web 05/25/2006
© 2006 American Chemical Society
of linear additivity places these classical issues in the contemporary context of quantitating the molecular switches of biologys posttranslational modifications (PTMs). For analysis of closely related peptides3 and intact proteins not using stable isotopes, few detailed studies have been described that dissect the various points of ratio biases during the analysis of a biological sample (depicted in Scheme 1 below). A lingering notion in large-molecule mass spectrometry has been that protein ions greater than ∼10 kDa may be less affected by biases B-E in Scheme 1 versus small peptides, because larger molecules change their physiochemical properties less for a given alteration in primary structure. Addressing bias C in Scheme 1, Gordon et al. recently demonstrated a quantitative relationship between the concentration of two proteins that were 97% identical and their FTMS peak ratios after electrospray ionization (ESI).4 They found a linear response in concentration with FTMS peak intensity over ∼1.5 orders of magnitude. Tandem MS (MS/MS) was not performed on these protein mixtures to examine whether fragment ion intensities correspond to intact ion intensities. Freitas5 and our group6 have claimed “semiquantitative” ratio determination of differentially modified forms of histone H4 that correlate well with the cellular perturbation employed (i.e., deacetylase inhibitors). Our group has also reported a qualitative correlation between abundance of precursors and abundance of their fragment ions when dissociating mixtures of two to eight intact proteins for multiplexed top down MS/MS.7,8 Electron capture dissociation (ECD) is well-suited for studying PTM-laden proteins such as eukaryotic histones6,9 because of its ability to generate fragment ion-rich data sets that allow nearly (3) Steen, H.; Jebanathirajah, J. A.; Springer, M.; Kirschner, M. W. Proc. Natl. Acad. Sci. U.S.A. 2005, 102 (11), 3948-53. (4) Gordon, E. F.; Mansoori, B. A.; Carroll, C. F.; Muddiman, D. C. J. Mass Spectrom. 1999, 34 (10), 1055-62. (5) Zhang, L.; Freitas, M. A.; Wickham, J.; Parthun, M. R.; Klisovic, M. I.; Marcucci, G.; Byrd, J. C. J. Am. Soc. Mass Spectrom. 2004, 15 (1), 77-86. (6) Pesavento, J. J.; Kim, Y. B.; Taylor, G. K.; Kelleher, N. L. J. Am. Chem. Soc. 2004, 126 (11), 3386-7. (7) Meng, F.; Cargile, B. J.; Patrie, S. M.; Johnson, J. R.; McLoughlin, S. M.; Kelleher, N. L. Anal. Chem. 2002, 74 (13), 2923-9. (8) Johnson, J. R.; Meng, F.; Forbes, A. J.; Cargile, B. J.; Kelleher, N. L. Electrophoresis 2002, 23 (18), 3217-23.
Analytical Chemistry, Vol. 78, No. 13, July 1, 2006 4271
Scheme 1
complete sequencing of small proteins de novo.10 To date, no one has studied carefully whether MS/MS (either threshold or electron-based dissociation methods) generates fragment ion ratios that accurately report on modification occupancies at specific sites. To the extent this is true, MS/MS of protein mixtures would not only serve to confirm their intact ratios but also increase the sophistication of MS/MS spectral interpretation. For example, if a protein is monoacetylated at one of two possible residues, MS would not be informative since each acetylated form is isomeric. However, the MS/MS fragment ions would differentiate the two and provide information on the relative amount of acetylation on each residue. ECD fragmentation of Tetrahymena H2B has detected the presence of particular subtypes,9 and more recently, a more quantitative and gene-specific analysis has been applied to human H2B,11 H2A,12 and H3.13 In this study, we report that solution ratios of defined mixtures of unmodified and acetylated H4 peptides (2 kDa) vary dramatically from the observed MS ratios, while solution ratios of recombinant H4 protein mutants (11 kDa) correspond closely with MS ratios. Regardless of this solution versus MS ion ratio discrepancy, ratios of fragment ion intensities (so-called FIRRs) correspond quantitatively to relative ratios of intact ion intensities for mixtures of synthetic histone H4 peptides with defined acetylation sites. In addition, we show that a ∼1:1 mixture of triacetylation isomers generates ∼1:1 FIRRs with a precision of ∼5%. By extending this quantitative MS/MS procedure to naturally occurring mixtures of histone H4 acetylation isomers (11 kDa), we are able to identify the sites of acetylation and provide a quantitative determination of the percent occupancy in mixtures with up to 15 distinct protein formssall biologically relevant. MATERIALS AND METHODS Synthetic Peptide and Recombinant H4 Protein Preparations. Six synthetic peptides representing the first 20 N-terminal amino acids of human histone H4 were a generous gift from Pat Nakatani. One peptide was unmodified, one was tetraacetylated (aK5/aK8/aK12/aK16), and four were defined PTM isomers of triacetylated forms (aK8/aK12/aK16, aK5/aK12/aK16, aK5/aK8/ aK16, aK5/aK8/aK12). After chemical synthesis, the amount of these peptides was individually quantified by amino acid analysis and was nominally in ∼30 µg/µL range. Working stocks of each peptide were prepared by diluting solutions down to a concentra(9) Medzihradszky, K. F.; Zhang, X.; Chalkley, R. J.; Guan, S.; McFarland, M. A.; Chalmers, M. J.; Marshall, A. G.; Diaz, R. L.; Allis, C. D.; Burlingame, A. L. Mol. Cell. Proteomics 2004, 3 (9), 872-86. (10) Zubarev, R. A.; Horn, D. M.; Fridriksson, E. K.; Kelleher, N. L.; Kruger, N. A.; Lewis, M. A.; Carpenter, B. K.; McLafferty, F. W. Anal. Chem. 2000, 72 (3), 563-73. (11) Siuti, N.; Roth, M. J.; Mizzen, C. A.; Kelleher, N. L.; Pesavento, J. J. J. Proteome Res. 2006, 5 (2), 233-9. (12) Boyne II, M. T.; Pesavento, J. J.; Mizzen, C. A.; Kelleher, N. L. J. Proteome Res. 2006, 5 (2), 248-53. (13) Thomas, C. E.; Kelleher, N. L.; Mizzen, C. A. J. Proteome Res. 2006, 5 (2), 240-7.
4272
Analytical Chemistry, Vol. 78, No. 13, July 1, 2006
tion of 1 µg/µL in methanol, water, and formic acid (49:49:2). These working stocks were then reanalyzed by HPLC as an additional verification/quantitation of their relative concentrations (vide infra). Xenopus laevis H4 was expressed in Escherichia coli using a plasmid kindly provided by Luger as described previously.14 H18A and H75A mutants were created using the QuickChange Mutagenesis kit (Stratagene). H4 solutions were quantitated by comparing signal intensities in SDS-PAGE and HPLC to that of a standard solution quantitated by amino acid analysis. HPLC. Prior to mass spectrometry, the concentrations of the working stocks of recombinant H4 and synthetic peptides to be compared were validated by assessing the peak areas obtained in reversed-phase chromatography (RPLC) on a 2-mm-i.d. C18 column (Vydac) with detection at 214 nm. Two volumes, chosen to contain ∼2 and ∼6 µg based on the original amino acid analyses, were analyzed for the working solution of each of the six synthetic peptides. Only minor variations were observed for each peptide stock concentration (average standard deviation of 20% deviation from known solution ratios), while for intact proteins the effect of ionization efficiency on PIRRs appears far less dramatic (e5% deviation from solution ratio). Regardless of ionization efficiency effects, the PIRRs matched closely to the FIRRs (