C H A P T E R
E L E V E N
Quantitative Proteome Mapping of Nitrotyrosines Diana J. Bigelow* and Wei-Jun Qian† Contents 1. Introduction 2. Multidimensional LC-MS/MS Provides Large Data Sets for Identification of Nitrotyrosine-Modified Proteins 3. Retention of Complexity in Samples Prepared from Global Proteomic Analysis 4. Confident Identification of Nitrotyrosine-Containing Peptides 5. Comparative Quantitation of Nitrotyrosine-Modified Peptide/Proteins 6. Summary References
192 193 195 197 198 203 203
Abstract An essential first step in the understanding disease and environmental perturbations is the early and quantitative detection of the increased levels of the inflammatory marker nitrotyrosine, as compared with its endogenous levels within the tissue or cellular proteome. Thus, methods that successfully address a proteome-wide quantitation of nitrotyrosine and related oxidative modifications can provide early biomarkers of risk and progression of disease, as well as effective strategies for therapy. Multidimensional separations LC coupled with tandem mass spectrometry (LC-MS/MS) has, in recent years, significantly expanded our knowledge of human (and mammalian model system) proteomes, including some nascent work in identification of posttranslational modifications. This chapter discusses the application of LC-MS/MS for quantitation and identification of nitrotyrosine-modified proteins within the context of complex protein mixtures presented in mammalian proteomes.
Cell Biology and Biochemistry Group, Division of Biological Sciences, Pacific Northwest National Laboratory, Richland, Washington Division of Biological Sciences, Environmental Molecular Sciences Laboratory, Pacific Northwest National Laboratory, Richland, Washington
Methods in Enzymology, Volume 440 ISSN 0076-6879, DOI: 10.1016/S0076-6879(07)00811-7
2008 Elsevier Inc. All rights reserved.
Diana J. Bigelow and Wei-Jun Qian
1. Introduction Recent advances in the sensitivity of mass spectrometry instrumentation, coupled with improved separations methods, now offer increasing opportunities for understanding cell and tissue responses at a whole proteome level. Current technologies allow the abundance of thousands of proteins to be monitored simultaneously. Thus, the entire proteome of simple organisms can be resolved, and for larger mammalian proteomes, these capabilities offer significant access into the changing protein landscape of cells and tissues in response to stress and disease, as well as a means to monitor the efficacy of therapeutic interventions. Moreover, sensitive identification of changes in both abundance and posttranslational modification of proteins can provide early and valuable biomarkers of disease, as well as insights into molecular mechanisms of disease progression. Despite the important role of posttranslational modifications in the activation of signal transduction pathways or as markers of inflammatory events, proteomics strategies to address this dynamic aspect of the proteome have not been fully exploited. In particular, in view of the frequent contribution of inflammation and oxidative stress in pathological tissue changes, 3-nitrotyrosine modifications are of particular interest in monitoring initiation and progression of disease. Indeed, since an antinitrotyrosine antibody became available over a decade ago, increased protein nitration has been reported in over 100 human pathologies and their animal or cell models (Pacher et al., 2007; Ye et al., 1996). However, associated information has lagged regarding the identity of nitrated proteins in specific pathophysiological conditions, the extent of their nitration, and how their modification alters protein and cellular function. Thus, the picture of pathways that explain the correlations between nitrotyrosine and disease is substantially incomplete. As an additional confounding factor, the cell death that often accompanies chronic pathology is likely to contribute to a limited understanding of mechanisms related to disease progression. Thus, detection of proteome changes at early stages in disease progression is essential in understanding disease and in the development of effective therapeutic strategies; an initial step requires the establishment of a baseline of endogenous oxidative stress from a quantitative identification of nitrotyrosine-modified proteins in tissues in the absence of pathology. From such analyses, additional information can be gained regarding the cellular distribution of nitrated proteins as a signature of reactive nitrogen chemistries in the cell, as well as potential vulnerabilities to inflammation and redox regulation of specific proteins and pathways; moreover, examination of nitrotyrosine sites suggests protein structural features that enhance nitration in vivo.
2. Multidimensional LC-MS/MS Provides Large Data Sets for Identification of Nitrotyrosine-Modified Proteins In order to establish a quantitative baseline level of protein nitration, several criteria must be met, i.e., (i) samples must be sufficiently complex so that their composition reflects that of the original cell or tissue; (ii) analytical techniques must be sensitive enough to provide the large data sets necessary to identify relatively low abundance modifications of multiple proteins; and (iii) methods must include a means to identify precise sites of modification sites. In large part, multidimensional LC or other high resolution methods for separation of peptides coupled with tandem mass spectrometry (LC/LC-MS/MS) have provided the best solution to these requirements. Typically, these methods from global proteomic analyses utilize complex protein samples of interest, e.g., cellular, tissue, or organelle homogenates, subjected to proteolytic digestion. The resulting peptides are separated according to physical properties, e.g., charge, size, or hydrophobicity, followed by a second dimension of separation, often inline with very sensitive mass spectrometry (MS). For each MS run (corresponding to one fraction from second-dimension separation), the five most abundant parent masses are selected for further fragmentation for MS/MS spectra from which peptide sequence information is obtained. Because most separation schemes provide more high-quality parent masses per fraction than the MS/MS capacity of the associated spectrometer, undersampling remains a technical limitation to this analytic approach (Qian et al., 2006). Nevertheless, the size of the data sets provided by current LC/LC-MS/MS capabilities are sufficiently large to challenge the overall throughput; for example, a recent global proteomic analysis of brain resulted in 751,000 individual MS/MS spectra (Wang et al., 2006). Subsequently, these experimental MS/MS spectra are matched with theoretical mass spectra calculated based on sequences from a given protein database for an organism of choice by means of automated database-searching algorithms (e.g., SEQUEST, MASCOT, X!Tandem) that provide peptide sequence identification (Craig and Beavis, 2004; Perkins et al., 1999; Yates et al., 1995). An alternative method, which has been commonly used for the identification of nitrotyrosine-modified proteins, consists of MS identification of peptides extracted from proteins bands/spots on one- or twodimensional electrophoresis gels after in-gel proteolysis. This approach has several shortcomings for quantitative global analysis of nitrated proteins, related primarily to high losses of protein during peptide extraction (Aulak et al., 2001; Castegna et al., 2003; Kanski et al., 2005a). Typically, nitrated proteins identified from gel-based approaches have been limited to
Diana J. Bigelow and Wei-Jun Qian
abundant and soluble proteins, which are detectable by the nitrotyrosine antibody and for which constituent peptides can be eluted from twodimensional gels in sufficient amounts for MS identification (Aulak et al., 2001; Castegna et al., 2003; Kanski et al., 2005a,b; Turko et al., 2003). Thus, in view of the common comigration of multiple proteins on electrophoresis gels and the incomplete sequence coverage of extracted proteins, considerable uncertainty exists regarding the identity of a nitrated peptide if, as is frequently the case, the nitrated sequence is not among the extracted peptides from a gel spot. Moreover, LC-MS/MS approaches avoid biases introduced through the use of antibodies for either enrichment or detection of nitrated proteins, as LC-MS/MS identification of nitrated peptides is based on the appearance of nitrotyrosine sites within a specific peptide sequence. Of note, several studies have shown substantial differential sensitivity to various proteins of antinitrotyrosine antibodies from different sources, differences that could lead to failure to detect major nitrated proteins (Barreiro et al., 2002; Sacksteder et al., 2006). Moreover, membrane proteins, which are typically lost by aggregation during the isoelectric focusing step of two-dimensional gels, are identified more abundantly when separations involve peptides rather than proteins (Wu and Yates, 2003). For example, membrane proteins comprised 26% of the proteins identified in a global screen of mouse brain, consistent with predictions that 20 to 30% of all open reading frames encode for membrane proteins (Ahram and Springer, 2004; Krogh et al., 2001; Wallin and von Heijne, 1998). However, it should be noted that transmembrane sequences themselves are in relatively low abundance in most LC-MS/MS proteomics screens (Blonder et al., 2002). In view of model studies that have shown more frequent nitration of intramembrane tyrosine probes as compared with tyrosines in aqueous solution, it might be expected that numerous nitrotyrosines present in vivo are being missed by current proteomics screens (Zhang et al., 2001). Thus, development of methods that enhance the recovery of hydrophobic membrane-spanning sequences will be important improvements for a more complete understanding of protein nitration in vivo. Other advantages of this analytical approach include its sensitivity and dynamic range, which provides identification of both high and low abundance cellular peptides and proteins. Finally, utilizing the same complex protein mixture to identify both nitrated peptides and their nonnitrated analogs makes a semiquantitative estimate of nitrotyrosine stoichiometry possible. Thus, this chapter focuses on the discussion of current applications of LC-MS/MS for quantitative mapping of nitrotyrosine sites within cells and tissues. In view of several excellent reviews available regarding developments in high-resolution mass spectrometric instrumentation and separation capabilities for proteomics, this chapter focuses on specific experimental strategies related to sample requirements, confident peptide identification, and quantitation of nitrotyrosine-modified peptide/proteins
(Domon and Aebersold, 2006; Ong and Mann, 2005; Qian et al., 2006; Sadygov et al., 2004; Shen and Smith, 2005; Smith et al., 2006).
3. Retention of Complexity in Samples Prepared from Global Proteomic Analysis As in any cost- and labor-intensive analyses, such as proteomics, sample quality is the most important factor. In order that nitroproteomic analyses are informative of the tissue, organelle, or cell of interest, samples should reflect the original protein composition. Thus, although several chemical and immunoaffinity methods have been developed to selectively enrich nitrated proteins for proteomic analysis, these approaches lack information regarding the extent of nitration within the original cell or tissue (Zhan and Desiderio, 2004; Zhang et al., 2007). A sample preparation that retains as much native complexity as is practical and is still compatible with its application to separation matrices is an essential first step in any proteomic analysis; in particular, this principle applies to the quantitative analysis of nitrotyrosines for which quantitation in context of the total protein is sought. Tissue homogenates that include the removal of connective tissue and other large particles (‘‘cell debris’’) are commonly used as a source of peptides for global analyses. However, recent proteomics studies have included a limited number of additional fractionation steps prior to highsensitivity separations that provide subproteomes to further reduce sample complexity, thus enhancing the detection of low abundance proteins and overall proteomic coverage. For example, Wang and co-workers (2006) subjected a portion of tryptic peptides from brain homogenate to cysteine-peptidyl enrichment (thiopropyl Sepharose) prior to cysteine alkylation; the remaining portion was analyzed without cysteine alkylation, essentially depleting this latter portion of cysteinyl-peptides to provide complementary proteomic samples. From the combined analysis of these two samples, from 1564 to 1859 additional proteins were identified as compared with the yield from either one of these samples. Similarly, the proteome coverage of human mammary epithelial cells was increased from 10 to 34% with cysteine-peptidyl enrichment (Liu et al., 2005). Other simple fractionation steps prior to proteolysis have been employed that permit the identification of more proteins from the combined results of each subproteome as compared with a single global fraction, such as a tissue homogenate (Foster et al., 2006; Wu et al., 2007). Figure 11.1 shows a comparison of results from two different protein fractionation strategies applied to mouse heart homogenates prior to LC-MS/MS analyses. One of these involves a single centrifugation step of the tissue homogenate that
Diana J. Bigelow and Wei-Jun Qian
3298 unique proteins
3432 unique proteins
Figure 11.1 Fractionation of mouse heart homogenates reduces sample complexity and increases proteome coverage.Venn diagrams indicate protein identifications resulting from different fractionation strategies applied to LC-MS/MS proteomic analyses of mouse hearts. (Left) Heart homogenates were centrifuged at 48,000 g; the resulting membrane pellets were solubilized with either 2% CHAPS or 30% TFE prior to processing for LC-MS/MS analysis in parallel to that for the supernatant fraction (Soluble). Combined analysis resulted in the identification of 3298 unique proteins; the total protein identified for each fraction is indicated in parentheses. Numbers inside the circles indicate identified proteins that are unique or in common with one or two other fractions. (Right) Heart homogenates were differentially centrifuged at 15,000 g, resulting in a mitochondrially enriched pellet (Mito), and at 48,000 g, resulting in an sarcoplasmic reticulum-enriched pellet (SR); and the 48,000 g supernatant was designated as a cytosolic-enriched (Cyt) fraction. Samples were digested and processed for LC-MS/MS analysis, resulting in the identification of 3432 unique proteins from the combined results; proteins unique or common with one or two other fractions are indicated within the circles as indicated for each fraction. For proteomic analysis, strong cation exchange was used for the first dimension of peptide separation, followed by separation of the resulting 30 fractions on a reversed-phase (C18) capillary HPLC system coupled online with an LTQ ion trap mass spectrometer (ThermoFinnigan San Jose, CA) using an in-house manufactured electrospray ionization interface. MS/MS data were searched against the mouse International Protein Index (IPI) database (version 1.25, available online at http://www.ebi.ac.uk/IPI) and a sequence-reversed IPI database (to assess false positives) using SEQUEST (ThermoFinnigan).
largely separates membrane vesicles from soluble supernatant proteins; the resulting membrane pellet is solubilized with either the detergent 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS) or the organic solvent trifluoroethanol (TFE). In particular, the use of two different solubilization agents permits enhanced numbers of unique proteins identified in the membrane fractions so that their total number approximates that of the soluble fraction. As an alternative method, differential
centrifugation was used to prepare fractions enriched in mitochondria, sarcoplasmic reticulum membranes, and cytosolic proteins (Fig. 11.1). This approach, too, increased the total proteins identified relative to either fraction alone or as compared with a single homogenate fraction, in which approximately 900 proteins were initially identified (data not shown). Sequence coverage was also improved. Similar to overall protein identification, nitrotyrosine-modified protein identification was also enhanced by these initial fractionation strategies. Any number of fractionation schemes might be devised with several caveats, i.e, (i) all subproteomes should be analyzed, avoiding the discarding of fractions, so that the combined results reflect the original proteome of interest and that (ii) within the LC-MS capabilities, fractionation/separations resolution increases have to be balanced with their concomitant decreases in overall sample throughput. In particular, the number of subproteome fractionations at the front end of global analysis increases total MS runs rapidly and therefore decreases throughput. For example, in the case when each experimental sample includes approximately 30 first-dimension separation fractions applied to the second-dimension separation online with the mass spectrometer, the three subproteome samples used in Fig. 11.1 require 90 MS runs per experimental sample, an analysis that requires significant instrumental and computational time under current system capabilities.
4. Confident Identification of Nitrotyrosine-Containing Peptides The automated database searching used to identify native peptides can be adjusted to accommodate searches for specific posttranslational modifications of interest by including the addition of an appropriate mass to the corresponding amino acid within peptide sequences. For example, in the case of nitrotyrosine, searches are performed for the addition of 44.98 Da applied to tyrosines within peptide sequences. Searches for additional modifications may be desired to widen the search for nitrated peptides, as, for example, cysteine carboxamidomethylation is a usual step in sample processing for MS analysis, requiring searches for 57.02 Da applied to cysteines. In a recent LC-MS proteomic analysis, almost two-thirds of nitrated peptides also contained at least one Cys residue; these are peptides that would elude identification without consideration of appropriate modifications on cysteines. Additionally, methionine sulfoxides (with their additional mass of 15.99 Da) should be included in searches for nitrated peptides, as these methionine modifications commonly result from nitrating chemistries, such as peroxynitrite, and are frequently present within nitrated peptides. In biological samples where oxidative stress and inflammation are expected,
Diana J. Bigelow and Wei-Jun Qian
other stable oxidative modifications may be informative and necessary for the efficient identification of nitrotyrosine modifications, as, for example, sulfinic (RSO2H) and sulfonic (RSO3H) acid derivatives of cysteines, methionine sulfones (MetSO2), nitrotryptophans, and bromo- and chlorotyrosines. However, the substantial computational times required for simultaneous searching of multiple modifications within large data sets place practical limits on most global proteomic searches to no more than three or four modifications at a time. The confidence of peptide identifications by automated database searching is based on the use of stringent criteria for filtering data. To address possible false-positive peptide identifications, sequence-reversed protein database searches have been used in order to establish filtering criteria that ensure low levels (e.g., <2%) of false-positive peptide identifications. Moreover, especially in the case of low abundance modifications such as nitrotyrosines, nitrated peptide identifications can be validated by manual inspection of the corresponding MS/MS spectra to ensure that (i) spectral peaks are clearly defined above the spectral noise; (ii) that no major unidentified peaks are present, and that (iii) major peak assignments are consistent with predicted patterns as illustrated by MS/MS spectra for the nitrated and unmodified versions of the peptide, DSYVAIANACCAPR (Fig. 11.2). As observed by comparison of these two spectra, all detected y fragment ions match for the two spectra, but all b ions containing the nitrotyrosine from the nitrated peptide (b3–b12) have masses that are 44.98 Da higher than the corresponding b ions from the nonnitrated peptide.
5. Comparative Quantitation of Nitrotyrosine-Modified Peptide/Proteins Pathophysiological or environmental perturbations often alter the protein abundances, including posttranslational modifications such as nitrotyrosines in cells, tissues, or biological fluids. The ability to quantitatively measure relative protein abundance and their modification differences between different conditions is essential for identifying key molecular targets or candidate biomarkers involved in different diseases. Current quantitative characterization of relative protein abundance differences can be categorized as two general approaches: (1) MS peptide ion intensitybased quantitation, often in combination with stable isotope labeling, and (2) MS/MS spectrum count (or peptide hit)-based quantitation. In the second approach, quantitation information is derived from the number of MS/MS sampling identifying a given protein within a complex mixture. Spectrum counts correspond to the number of MS/MS spectra observed for each distinct peptide, which may include repetitive identification or the
R.DSY#VAIANAC!C!APR.F Xcorr = 3.74 y5 2+ ion m/z = 807.02
b6 y6 b3
b12 b11 1341.48
y4 b4 b5
R.DSYVAIANAC!C!APR.F y7 y8 Xcorr = 4.11 2+ ion y5 m/z = 784.80
b7 y9 y10
600 800 1000 1200 m/z(daltons/unit of charge)
Figure 11.2 MS/MS spectra of the nitrated (A) and nonnitrated (B) peptide, DSYVAIANACCAPR, from the vesicular inhibitory amino acid transporter. Y# indicates nitrotyrosine modification, and C! indicates a cysteinyl residue modified by carboxamidomethylation.The two spectra match well on all detected b and y fragment ions, with the exception that b ions containing nitrotyrosine, i.e., b3^b12, from the nitrated peptide have masses 44.98 Da higher than the corresponding b ions from the nonnitrated peptide, as illustrated by arrows for b12.
same peptide in one or multiple first-dimension fractions. These values have been demonstrated to provide a reproducible and semiquantitative measure of peptide concentration in complex samples (Qian et al., 2005).
Diana J. Bigelow and Wei-Jun Qian
Because nitrotyrosine is often measured within the context of the global proteome, both approaches can theoretically be applied for quantitative measurements of differences in nitration levels. However, the use of stable isotope labeling for the quantitation of nitrotyrosine modifications has not been reported to date. As discussed here, MS/MS spectrum counts derived from current LC-MS/MS capabilities provide the most reliable quantitation of nitrotyrosine modification extent. For MS ion intensity-based quantitation, most approaches have involved the use of stable isotope labeling to achieve good accuracy in quantifying the relative abundance differences. Many stable isotope-labeling methods have been introduced successfully for quantitative proteomics and are reviewed elsewhere (Ong and Mann, 2005; Qian et al., 2004). In general, stable isotope labels can be introduced into proteins or peptides by metabolic labeling, chemical labeling on specific functional groups, and enzymatic transfer of 18O from water to the C terminus of peptides. All of these methods incorporate either a light or a heavy version of the stable isotope into chemically identical peptides from two different samples, and the resulting peptide pairs are differentiated by MS due to the mass difference between the light and the heavy isotope-coded pair. The ratio of MS ion intensities for the peptide pair can then accurately indicate the abundance ratios of the peptide/protein from the two different samples. More recently, there has been a significant interest in applying ‘‘label-free’’ direct quantitation due to the greater flexibility for comparative analyses and simpler sample processing procedures compared to labeling approaches (Qian et al., 2006; Wang et al., 2003). The isotope labeling and label-free approaches are complementary, and each approach has different sources of variations. While the stable isotope-labeling approach is generally more accurate in quantitation as compared with the label-free approach as a consequence of the unbiased measurement of peptide pairs by MS and the minimum variations from sample processing postlabeling, isotopic ratios often have a limited dynamic range for quantitating large changes as a consequence of the overlap of isotopic envelopes between light and heavy members of the pair. However, the label-free approach is often more ideal for the quantitation of more dramatic changes in protein abundances or in estimating the stoichiometry of posttranslational modification of low abundance modifications as nitrotyrosines. With the proper use of replicate analyses and data normalization to mitigate changes in instrumental performance between analyses, accurate comparative quantitation can be achieved for biological and clinical applications. The MS/MS spectrum counting approach has been reported as an alternative approach for relative quantitation based on the correlation of protein abundances with the number of MS/MS spectra (or peptide hits) identifying a given proteins (Liu et al., 2004; Qian et al., 2005; Zybailov et al., 2005). It should be noted that the spectrum count approach is a semiquantitative method for comparative analyses because the spectrum
counts may not have a linear correlation with protein abundances. Additionally, the spectrum count approach will become problematic when proteins are observed with only a few counts in both conditions. Despite these potential disadvantages, this approach has been increasingly applied for several reasons, including (1) the experiments can be performed on a common ion-trap instrument without the need of high-resolution MS measurements, (2) the approach can be easily coupled with multiple dimensional LC separations, and (3) the reproducibility in terms of spectrum counts is observed for both one-dimensional LC (Liu et al., 2004, 2006) and two-dimensional LC-MS/MS (Qian et al., 2005). Moreover, the spectrum count information provides an estimation of the relative abundances of different proteins within the same sample (Wang et al., 2006). Specifically, when protein modifications such as nitrotyrosine are of interest, the extent of modification or the stoichiometry of modifications can be estimated by comparing spectrum counts of the modified peptides to total peptides from the global proteome for a given proteins. Here, the large dynamic range of spectrum counts is valuable, as typical levels of nitrated peptides relative to their nonnitrated analogs are quite low, <1/1000, levels that would disallow reliable quantitation from isotopic ratios. Another limitation of using stable isotope ratios for the quantitation of posttranslational modifications is that the number of peptide pairs resolved by this method is generally considerably smaller than the peptides identified by label-free approaches, which would further reduce the number of detected nitrotyrosine-modified peptides that are quantifiable. A more common problem for spectrum count-associated quantitation of nitrotyrosines is the detection either of only a few counts of nitrated peptides or of differences derived from different experimental conditions. These low levels may result from incomplete sampling because of stringent criteria for identification and their low abundance; guidelines for the evaluation and statistical significance of spectrum counts at low levels have been discussed previously (Qian et al., 2005). Reproducibility for LC-MS/MS proteomic analyses can be excellent; a recent study comparing results of proteomic analyses of nine individually processed technical replicates of human plasma samples showed very low variation in peptide intensities with an overall Pearson’s correlation coefficient of 0.94 0.02 (Qian et al., 2006). In comparison, plasma samples from nine individual human patients exhibited significantly more variation with a correlation coefficient of 0.85 0.06; however, analyses from nine individual mouse plasma samples showed only a slightly reduced correlation (0.92 0.05) as compared with technical replicates. Studies such as these highlight the importance of minimizing any instrumental contribution to the variability in proteomics results in order to enhance the detection of experimental differences.
Diana J. Bigelow and Wei-Jun Qian
As an illustration of spectrum counts used to quantitate nitrotyrosine levels in several tissues, we compared the results of three multidimensional LC-MS/MS proteomic screens of mouse brain, heart, and skeletal muscle, identifying endogenous nitrotyrosine-modified peptides and associated proteins from SEQUEST searches that consider methionine sulfoxides and carboxymethylated cysteines as well as nitrotyrosines. Calculating the ratio of the sum of spectrum counts associated with all nitrotyrosinecontaining peptides relative to the sum of all tyrosine-containing peptides, we found significantly different levels of endogenous protein nitration in these three tissues that correlate well with the differential staining intensities exhibited by nitrotyrosine immunoblots of these samples where the greatest abundance of nitrated proteins is found in skeletal muscle followed by heart followed by brain (Fig. 11.3). Notably, the nitrotyrosine immunoblot of A
Sk muscle 3
220 160 120 100 80
98 62 49 38
40 14 30 6 − B NitroTyr/Tyr (mol%):
Figure 11.3 Different levels of protein nitration in heart, skeletal (Sk) muscle, and brain. Homogenate proteins from heart, hind limb skeletal muscle, and brain were separated on SDS-PAGE prior to immunoblotting with the antinitrotyrosine antibody (A) for comparison with relative levels of nitrotyrosine-containing peptides identified from multidimensional LC-MS/MS analyses of parallel samples (B). Immunoblots show heart and skeletal muscle homogenates from three individual animals; shown are immunoblots for normal () and for (þ) MPTP-lesioned brains, prepared as described previously (Sacksteder et al., 2006). Proteomic-derived values for nitrotyrosine levels were determined from the ratios of (i) the sum of spectrum counts associated with identified nitrotyrosines in peptides to (ii) the sum of total spectrum counts for identified tyrosines in peptides. Proteomic analyses were performed essentially as described in Fig. 11.1 and detailed by Sacksteder and co-workers (2006), with the exception that muscle homogenates were prefractionated according to the scheme outlined in the left side of Fig.11.1.
homogenate from the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-lesioned brain, a neurotoxin-induced model of Parkinson’s disease, exhibits more intense staining relative to normal brain of the same major nitrated bands, highlighting the importance of identifying nitrationsensitive proteins as potential indicators of neurodegeneration and pathology. Indeed, from LC-MS/MS identification of endogenously nitrated proteins in normal brain, more than half of the identified nitrated proteins have been previously reported to be associated with neurodegeneration, further supporting the link between nitration-sensitive proteins and their vulnerability to disease.
6. Summary Multidimensional LC MS/MS offers expanding possibilities in the quantitative mapping of the nitroproteome as a signature of inflammatory events in normal and disease-progressing tissues. Recent global proteomics have begun to identify significantly greater numbers of nitrated proteins within the context of the complex mixture, as well as to define endogenous levels of inflammation in different tissues. The detection of multiple nitrotyrosine-modified proteins in tissues is just the tip of the iceberg with regard to posttranslational modifications of the proteome. Future work will be aimed toward greater coverage of these modified proteins.
REFERENCES Ahram, M., and Springer, D. L. (2004). Large-scale proteomic analysis of membrane proteins. Expert Rev. Proteomics 1, 293–302. Aulak, K. S., Miyagi, M., Yan, L., West, K. A., Massillon, D., Crabb, J. W., and Stuehr, D. J. (2001). Proteomic method identifies proteins nitrated in vivo during inflammatory challenge. Proc. Natl. Acad. Sci. USA 98, 12056–12061. Barreiro, E., Comtois, A. S., Gea, J., Laubach, V. E., and Hussain, S. N. (2002). Protein tyrosine nitration in the ventilatory muscles: Role of nitric oxide synthases. Am. J. Respir. Cell. Mol. Biol. 26, 438–446. Blonder, J., Goshe, M. B., Moore, R. J., Pasa-Tolic, L., Masselon, C. D., Lipton, M. S., and Smith, R. D. (2002). Enrichment of integral membrane proteins for proteomic analysis using liquid chromatography-tandem mass spectrometry. J. Proteome Res. 1, 351–360. Castegna, A., Thongboonkerd, V., Klein, J. B., Lynn, B., Markesbery, W. R., and Butterfield, D. A. (2003). Proteomic identification of nitrated proteins in Alzheimer’s disease brain. J. Neurochem. 85, 1394–1401. Craig, R., and Beavis, R. C. (2004). TANDEM: Matching proteins with tandem mass spectra. Bioinformatics 20, 1466–1467. Domon, B., and Aebersold, R. (2006). Mass spectrometry and protein analysis. Science 312, 212–217. Foster, L. J., de Hoog, C. L., Zhang, Y., Zhang, Y., Xie, X., Mootha, V. K., and Mann, M. (2006). A mammalian organelle map by protein correlation profiling. Cell 125, 187–199.
Diana J. Bigelow and Wei-Jun Qian
Kanski, J., Behring, A., Pelling, J., and Schoneich, C. (2005a). Proteomic identification of 3-nitrotyrosine-containing rat cardiac proteins: Effects of biological aging. Am. J. Physiol. Heart Circ. Physiol. 288, H371–H381. Kanski, J., Hong, S. J., and Schoneich, C. (2005b). Proteomic analysis of protein nitration in aging skeletal muscle and identification of nitrotyrosine-containing sequences in vivo by nanoelectrospray ionization tandem mass spectrometry. J. Biol. Chem. 280, 24261–24266. Krogh, A., Larsson, B., von Heijne, G., and Sonnhammer, E. L. (2001). Predicting transmembrane protein topology with a hidden Markov model: Application to complete genomes. J. Mol. Biol. 305, 567–580. Liu, H., Sadygov, R. G., and Yates, J. R., 3rd (2004). A model for random sampling and estimation of relative protein abundance in shotgun proteomics. Anal. Chem. 76, 4193–4201. Liu, T., Qian, W. J., Chen, W. N., Jacobs, J. M., Moore, R. J., Anderson, D. J., Gritsenko, M. A., Monroe, M. E., Thrall, B. D., Camp, D. G., 2nd, and Smith, R. D. (2005). Improved proteome coverage by using high efficiency cysteinyl peptide enrichment: The human mammary epithelial cell proteome. Proteomics 5, 1263–1273. Liu, T., Qian, W. J., Mottaz, H. M., Gritsenko, M. A., Norbeck, A. D., Moore, R. J., Purvine, S. O., Camp, D. G., 2nd, and Smith, R. D. (2006). Evaluation of multiprotein immunoaffinity subtraction for plasma proteomics and candidate biomarker discovery using mass spectrometry. Mol. Cell Proteomics 5, 2167–2174. Ong, S. E., and Mann, M. (2005). Mass spectrometry-based proteomics turns quantitative. Nat. Chem. Biol. 1, 252–262. Pacher, P., Beckman, J. S., and Liaudet, L. (2007). Nitric oxide and peroxynitrite in health and disease. Physiol. Rev. 87, 315–424. Perkins, D. N., Pappin, D. J., Creasy, D. M., and Cottrell, J. S. (1999). Probability-based protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 20, 3551–3567. Qian, W. J., Camp, D. G., 2nd, and Smith, R. D. (2004). High-throughput proteomics using Fourier transform ion cyclotron resonance mass spectrometry. Expert Rev. Proteomics 1, 87–95. Qian, W. J., Jacobs, J. M., Camp, D. G., 2nd, Monroe, M. E., Moore, R. J., Gritsenko, M. A., Calvano, S. E., Lowry, S. F., Xiao, W., Moldawer, L. L., Davis, R. W., Tompkins, R. G., et al. (2005). Comparative proteome analyses of human plasma following in vivo lipopolysaccharide administration using multidimensional separations coupled with tandem mass spectrometry. Proteomics 5, 572–584. Qian, W. J., Jacobs, J. M., Liu, T., Camp, D. G., 2nd, and Smith, R. D. (2006). Advances and challenges in liquid chromatography-mass spectrometry-based proteomics profiling for clinical applications. Mol. Cell Proteomics 5, 1727–1744. Sacksteder, C. A., Qian, W. J., Knyushko, T. V., Wang, H., Chin, M. H., Lacan, G., Melega, W. P., Camp, D. G., 2nd, Smith, R. D., Smith, D. J., Squier, T.C, and Bigelow, D. J. (2006). Endogenously nitrated proteins in mouse brain: Links to neurodegenerative disease. Biochemistry 45, 8009–8022. Sadygov, R. G., Cociorva, D., and Yates, J. R., 3rd (2004). Large-scale database searching using tandem mass spectra: Looking up the answer in the back of the book. Nat. Methods 1, 195–202. Shen, Y., and Smith, R. D. (2005). Advanced nanoscale separations and mass spectrometry for sensitive high-throughput proteomics. Expert Rev. Proteomics 2, 431–447. Smith, R. D., Tang, K., and Shen, Y. (2006). Ultra-sensitive and quantitative characterization of proteomes. Mol. Biosyst. 2, 221–230. Turko, I. V., Li, L., Aulak, K. S., Stuehr, D. J., Chang, J. Y., and Murad, F. (2003). Protein tyrosine nitration in the mitochondria from diabetic mouse heart: Implications to dysfunctional mitochondria in diabetes. J. Biol. Chem. 278, 33972–33977.
Wallin, E., and von Heijne, G. (1998). Genome-wide analysis of integral membrane proteins from eubacterial, archaean, and eukaryotic organisms. Protein Sci. 7, 1029–1038. Wang, H., Qian, W. J., Chin, M. H., Petyuk, V. A., Barry, R. C., Liu, T., Gritsenko, M. A., Mottaz, H. M., Moore, R. J., Camp Ii, D. G., Khan, A. H., Smith, D. J., et al. (2006). Characterization of the mouse brain proteome using global proteomic analysis complemented with cysteinyl-peptide enrichment. J. Proteome Res. 5, 361–369. Wang, W., Zhou, H., Lin, H., Roy, S., Shaler, T. A., Hill, L. R., Norton, S., Kumar, P., Anderle, M., and Becker, C. H. (2003). Quantification of proteins and metabolites by mass spectrometry without isotope labeling or spiked standards. Anal. Chem. 75, 4818–4826. Wu, C. C., and Yates, J. R., 3rd (2003). The application of mass spectrometry to membrane proteomics. Nat. Biotechnol. 21, 262–267. Wu, W. W., Wang, G., Yu, M. J., Knepper, M. A., and Shen, R. F. (2007). Identification and quantification of basic and acidic proteins using solution-based two-dimensional protein fractionation and label-free or 18O-labeling mass spectrometry. J. Proteome Res. 6, 2447–2459. Yates, J. R., 3rd, Eng, J. K., and McCormack, A. L. (1995). Mining genomes: Correlating tandem mass spectra of modified and unmodified peptides to sequences in nucleotide databases. Anal. Chem. 67, 3202–3210. Ye, Y. Z., Strong, M., Huang, Z. Q., and Beckman, J. S. (1996). Antibodies that recognize nitrotyrosine. Methods Enzymol. 269, 201–209. Zhan, X., and Desiderio, D. M. (2004). The human pituitary nitroproteome: Detection of nitrotyrosyl-proteins with two-dimensional Western blotting, and amino acid sequence determination with mass spectrometry. Biochem. Biophys. Res. Commun. 325, 1180–1186. Zhang, H., Joseph, J., Feix, J., Hogg, N., and Kalyanaraman, B. (2001). Nitration and oxidation of a hydrophobic tyrosine probe by peroxynitrite in membranes: Comparison with nitration and oxidation of tyrosine by peroxynitrite in aqueous solution. Biochemistry 40, 7675–7686. Zhang, Q., Qian, W. J., Knyushko, T. V., Clauss, T. R., Purvine, S. O., Moore, R. J., Sacksteder, C. A., Chin, M. H., Smith, D. J., Camp, D. G., 2nd, Bigelow, D. J., and Smith, R. D. (2007). A method for selective enrichment and analysis of nitrotyrosinecontaining peptides in complex proteome samples. J. Proteome Res. 6, 2257–2268. Zybailov, B., Coleman, M. K., Florens, L., and Washburn, M. P. (2005). Correlation of relative abundance ratios derived from peptide ion chromatograms and spectrum counting for quantitative proteomic analysis using stable isotope labeling. Anal. Chem. 77, 6218–6224.