ISSN 10619348, Journal of Analytical Chemistry, 2010, Vol. 65, No. 13, pp. 1295–1310. © Pleiades Publishing, Ltd., 2010. Original Russian Text © E.D. Kan’shin, I.E. Nifant’ev, A.V. Pshezhetskii, 2010, published in Massspektrometria, 2010, 6(2), pp. 103–120.

REVIEWS

Mass Spectrometric Analysis of Protein Phosphorylation E. D. Kan’shina, b, I. E. Nifant’evb, and A. V. Pshezhetskiia, c, d a

University of Montreal, CHU SainteJustine, l175 Cote SteCatherine, Montreal (Quebec), H3T 1C5 Canada b Department of Chemistry, Moscow State University, Moscow, 119991 Russia cFaculty of Medicine, Department of Pediatrics and Biochemistry, University of Montreal, 3175 Cote SteCatherine, Montreal (Quebec), H3T 1C5 Canada dFaculty of Medicine, Department of Anatomy and Cell Biology, McGill University, 845 W. Sherbrooke St. W. Montreal (Quebec), H3A 2T5 Canada email: [email protected] Received September 15, 2008

Abstract—Phosphorylation is one of the most common posttranslational modifications of proteins in eukaryotic cells; it plays an important role in a wide spectrum of biological processes. This makes its study an important task for understanding cell functioning mechanisms. The aim of phosphoproteomics is a global mass spectral analysis of the phosphoprotein composition of cells, i.e., phosphoproteome. Nowadays, new effective methods are actively developed, which succeed not only in the detection of phosphorylated proteins but also in the determination of phosphorylated amino acid residues (phosphorylation sites) and in the quan titative comparison of phosphorylation among several specimens. Despite the analysis of protein phosphory lation remains a complicated problem, the available methods nowadays allow the detection of thousands of phosphorylation sites in the very same experiment. The present review covers the main methods utilized in contemporary phosphoproteomics: phosphoprotein and phosphopeptides enrichment as well as the mass spectrometric analysis of protein phosphorylation. Keywords: phosphoproteomics, mass spectrometry, phosphoprotein enrichment, peptide enrichment DOI: 10.1134/S1061934810130010

TABLE OF CONTENTS ABBREVIATIONS USED 1. PHOSPHORYLATION OF PROTEINS 2. SELECTIVE ENRICHMENT OF PHOS PHOPROTEINS AND PHOSPHOPEPTIDES —Use of antibodies —IMAC methods —MOAC methods —Cationexchange chromatography —Chemical modification of phosphonate groups 3. MASS SPECTROMETRY IN THE ANALYSIS OF PROTEIN PHOSPHORYLATION —Ionization of phosphopeptides —Alternative methods for the fragmentation of phosphopeptides —Use of multistep fragmentation MSn (n > 2) —Registration methods for product ions and the elimination of a neutral fragment —Use of enzymatic dephosphorylation 4. COMPARATIVE PHOSPHOPROTEOMICS —Semiquantitative methods —Quantitative methods —In vivo isotope modification of proteins

— Chemical modification of proteins and peptides by isotopic tags CONCLUSIONS REFERENCES ABBREVIATIONS USED IMAC—immobilized metal affinity chromatogra phy MOAC—metal oxide affinity chromatography SCX—strong cationexchange chromatography ICAT—isotopically coded affinity tags iTRAQ—isobaric tags for relative and absolute quantification (of proteins) HPLC—highperformance liquid chromatogra phy ESI—electrospray ionization MALDI—matrixassisted laser desorption/ioniza tion CID—collision induced dissociation, fragmenta tion due to collisions with neutral atoms ETD—electron transfer dissociation, fragmenta tion due to transfer of electron

1295

1296

KAN’SHIN et al.

ECD—electron capture ionization, fragmentation due to electron capture FTICR—fourier transform–ion cyclotron reso nance PAIs—protein abundant indices, coefficients of proteins’ abundance. SILAC—stable isotope labelling with amino acids in cell culture, insertion of tags due to addition of iso topicallylabelled amino acids into cell culture. PhIAT—phosphoprotein isotopecoded affinity tag. PhIST—phosphoprotein isotopecoded solid phase tag. 1. PHOSPHORYLATION OF PROTEINS The reverse phosphorylation of proteins appears one of the basic mechanisms for the regulation of their activity and affects a wide spectrum of processes within living cells, including signaling, differentiation, proliferation, and the control of cell cycle and meta bolism. This makes the analysis of phosphoproteins an important task in studying key cell functions. It is commonly established that around 30% of the proteins can be phosphorylated in one moment or another [1], and the number of sites for protein phosphorylation in a human proteome is estimated at a level of 100000 [1]. The results of the sequenation of human genome also support the physiological importance of phospho rylation: genes, which are coding enzymes which, in their turn, are responsible for the phosphorylation and dephosphorylation of proteins, correspond to 2% of the human genome at least [2]. The phosphorylation of amino acid residues of serine, threonine, and tyrosine is the most abundant [3]. In the recent study of HeLa cells, it was shown that the ratio between the peptides phosphorylated on serine, threonine, and tyrosine made up 86.4, 11.8, and 1.8%, respectively [4]. Phosphorylation of amino (arginine and lysine), carboxyl (asparaginic and glutamic acids), and thiol groups (cysteine) is also observed, but it is less abundant and also characterized by low stability. Various approaches are known to provide informa tion about phosphoproteins. The most common among them is represented by proteomics, the science investigating the full entirety of phosphoproteins in a cell or organism, i.e., phosphoproteome. Because of the high complexity of protein composition in biolog ical systems, the analysis of both proteome and phos phoproteome represents a difficult task. To solve this task, mass spectrometry is finding increasing wide application [5]; it is combined with various techniques of protein and peptide separation [6], namely, gel electrophoresis [7], HPLC [8], capillary electrophore sis [9, 10], etc. One of the most abundant approaches used in phosphoproteomics (“bottomup” analysis) utilizes the mass spectrometric analysis of phospho

peptides obtained by the enzymatic cleavage of protein mixtures. In this approach, depending on the com plexity of the initial sample, one selects a method for their separation at the level of proteins or peptides. A sequential combination of several separation tech niques (twodimensional, threedimensional, etc., separation) is often used. The approach based on the analysis of intact phosphoproteins (“topdown” anal ysis) is less common. The mass spectrometric analysis of phosphoryla tion in biological objects is a more complicated task than the analysis of nonphosphorylated proteins for the following reasons: 1. Phosphoproteins form only a minor part of the total cell protein composition (usually no more than 10%); moreover, the phosphorylation of amino acid residues in proteins is often incomplete and can make up a few percent only [11, 12]; 2. In phosphoproteome analysis, in addition to the determination of the amino acid sequence of a phos phoprotein itself, it is necessary to determine the loca tion of the phosphorylated amino acid residue(s), i.e., the so called phosphorylation site; 3. The presence of nonphosphorylated peptides may suppress the ionization of phosphopeptides; 4. Often, the main direction of phosphopeptide fragmentation under the conditions of tandem mass spectrometry is represented by the elimination of phosphate, which results in poorly informative spectra and makes troublesome the determination of the pep tide amino acid sequence and the phosphorylation site; 5. The very same protein can be phosphorylated at several different positions, which leads to the possibil ity of different phosphorylated isoforms and the increased complexity of samples; 6. Phosphorylation is an unstable modification; in an experiment phosphoproteins and phosphopeptides may undergo dephosphorylation (for example, if the activity of cell phosphatases was retained [13]). However, despite the listed problems, the develop ment of increasingly sensitive instruments, novel methods for the fragmentation of phosphopeptides, as well as chromatography approaches to the selective enrichment of phosphoproteins and phosphopeptides allows researchers to define thousands phosphoryla tion sites within one experiment [14, 15]. The majority of approaches used for the analysis of phosphoproteome can be divided into two groups, namely, the selective enrichment of phosphoproteins and phosphopeptides, which makes it possible to get rid of the nonphosphorylated components of the sample, and various mass spectrometric tricks assisting in the more effective detection of phosphopeptides. The aim of the present review was to describe the basic methods used for the enrichment of phosphop roteins and phosphopeptides and the mass spectro

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION

1297

Protein mixture

Specific binding

Immunoprecipitation (pTyr) or IMAC (pSer/pThr/pTyr)

Antibody

Proteolysis

Protein

Phosphorylated proteins Proteolysis

IMAC (Fe3+/Ga3+/Zr4+/Al3+) MOAC (TiO2/ZrO2/Al2O3)

P P

Chemical modification

Peptides

Phosphorylated proteins

SCH

Phosphopeptides

P

βelimination/ phosphoamidates

Stationary phase

Mass spectrometric analysis

Fig. 1. Scheme illustrating the basic methods of phosphop rotein and phosphopeptide determination used in phos phoproteomics. For identification purpose, the proteins are first subjected to proteolythic cleavage to peptides. Meanwhile, there is an opportunity to perform the prelim inary isolation of phosphoproteins by immunoprecipita tion on antibodies specific to phosphotyrosine residues or using IMACtechniques. For the consequent isolation of phosphopeptides, IMAC and MOAC methods are used, and also the approaches based on the chemical modifica tion of phospho groups. The preliminary isolation of phos phopeptides is also performed by SCX.

metric analysis of phosphorylation as well as the meth ods for the quantitative comparison of phosphopro teomes (quantitative phosphoproteomics). 2. SELECTIVE ENRICHMENT OF PHOSPHOPROTEINS AND PHOSPHOPEPTIDES As mentioned above, the analysis of phosphopro teome is substantially complicated by the low concen tration of phosphoproteins and corresponding phos phopeptides in the samples. At the same time, the presence of nonphosphorylated peptides can sup press the ionization of phosphopeptides under the conditions of mass spectrometric analysis. Thus, the task of the analysis of protein phosphory lation (determination of phosphopeptides in the pres ence of nonphosphorylated peptides) can be com pared with the search for a needle in a haystack. Meanwhile, the function of a magnet is performed by the methods for the enrichment of phosphoproteins and phosphopeptides, which help us to concentrate on the exclusive analysis of a specific sample part (Fig. 1). Use of antibodies. The method under consideration is based on the use of antibodies specific to phospho rylated proteins and peptides. Immunoaffine chro matography, in which the antibodies are immobilized on a stationary phase and then used to capture the phosphorylated amino acid residues, has found the widest application. Therein the nonphosphorylated components of the sample can be removed, while phosphoproteins/phosphopeptides are eluted later for JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

Nonphosphorylated proteins

Fig. 2. Immunoprecipitation of phosphorylated pro teins/peptides. Antibodies are immobilized on a stationary phase. In the interaction with the phase, phosphorylated proteins are retained due to highly specific interaction with antibodies, while nonphosphorylated proteins are removed by washing. The phosphorylated proteins are eluted under denaturation conditions (1 M glycine buffer, pH 2.3) and used for subsequent analysis.

analysis (Fig. 2). An interesting approach is known, in which antibodies are used to recognize the specific amino acid motif for a specific kinase, and therefore discriminate all its phosphorylated substrates [16, 17]. The high specificity of the antigen–antibody inter action provides the major advantage of this group of methods. However, one needs to pay attention to the relatively high cost of antibodies and the unavailability of antibodies highly specific to phosphoserine and phosphothreonine [18]. Hence, timmunoprecipita tion finds the main application for the investigation of phosphorylation on tyrosine [19–21], while for the isolation of proteins phosphorylated on serine and threonine, one has to utilize expensive mixtures of antibodies, which react with different epitopes charac teristic for the phosphorylation of the respective amino acids [18]. The low concentration of proteins phosphorylated on tyrosine (generally below 0.05%) often requires the use of rather large amounts of sam ples. Thus, for example, in paper [22], using 80 mg of a protein sample and immunoprecipitation, the authors identified 385 different sites of phosphoryla tion on tyrosine. As a rule, immunoprecipitation is used for the enrichment of proteins but can be com bined with other methods for the preconcentration of phosphoproteins and phosphopeptides [23]. Therefore, for the time being, the use of antibodies is strongly recommended for the study of phosphory lation on tyrosine residues only and leaves room for other methods for the isolation of phosphoproteins and phosphopeptides. IMAC methods. One of most frequently used approaches for the enrichment of phosphoproteins and phosphopeptides appears to be affine chromatog raphy on phases with immobilized metals (IMAC). The method exploits the ability of some metals to form No. 13

2010

1298

KAN’SHIN et al.

Metal (Fe , Ga3+, Zr4+, Al3+) Peptide/ protein 3+

Stationary phase

Nonphosphorylated peptides/proteins

Chelating group (iminodiacetic/nitrile triacetic acids)

Eluate containing phosphopeptides/ phosphoproteins

Fig. 3. Isolation of phosphoproteins and phosphopeptides by IMAC. The metal ions (Fe3+, Ga3+, Zr4+, Al3+) are immobilized on a stationary phase with attached chelating groups (as a rule, iminodiacetic or nitrilotriacetic acids). The phospho groups of peptides/proteins form complexes with metals on the phase, which offers a tool for the pre concentration of phosphorylated peptides/proteins, which can then be eluted (solutions with high pH, EDTA) for subsequent analysis.

strong complexes with oxygencontaining ligands, such as the phospho group. We note that IMAC was first developed in 1975 for the isolation of proteins containing amino acid sequences of six or more histi dine residues [24]. At the same time, Ni2+ ions were used as a metal. For the enrichment of phosphopep tides, IMAC was first utilized in 1987 [25]. Ions Fe3+ and Ga3+ have found the widest application for the isolation of phosphoproteins and phosphopeptides [26]; however, ions Zr4+ and Al3+ are also used [27, 28]. As stationary phases, researchers generally use sepharose premodified by chelating groups (as a rule, iminodiacetic and nitrilotriacetic acids). The immobi lization of metals is attained due to complex forma tion. The principle of enrichment is based on the reten tion of phosphopeptides on a stationary phase due to the complexation of phospho groups with metals, while nonphosphorylated peptides can be removed by washing the column with the respective buffer solu tion. Phosphopeptides are then eluted either by solu tions with high pH (for instance, 0.1 M NH4HCO3 or NH4OH) or by a 0.1 M solution of EDTA (Fig. 3). The IMAC methods can be utilized for the enrich ment of peptides phosphorylated on serine, threonine, and tyrosine residues. IMAC is compatible with HPLC; at the same time, the eluted phosphopeptides can be applied directly on a reversedphase column the outlet of which is connected to a mass spectrome ter (with electrospray ionization). The listed points have made IMAC one of the most widely used approaches for the isolation of phosphopeptides for the time being. The main drawback of IMAC techniques deals with the unspecific binding of nonphosphorylated peptides with low values of pI. This refers to the fact that in addition to phospho groups, peptides also con

tain free carboxyl groups, which, being oxygencon taining donors, are able to bind with metals on the sur face of a stationary phase. The strength of this interac tion grows with the number of carboxyl groups, which leads to the coisolation of nonphosphorylated acidic peptides bearing several residues of glutamic and/or asparaginic acid. To increase selectivity, preconcentration is per formed in acidic media [29] in the presence of rela tively large amounts of organic solvents (for example, in the mixture acetonitrile : methanol : water = 1 : 1 : 1). The methylation of free carboxyl groups of peptides prior to the preconcentration is also performed [30]. The approach specified favors the substantial enhancement of the selectivity, in the pioneer paper on the study of the yeast S. cerevisiae phosphoproteome, the authors identified 383 phosphorylation sites. Moreover, the use of a mixture of CH3OH and CD3OH for methylation provides a way to insert stable isotopes and then perform comparative quantitative analysis of phosphopeptides [23]. The disadvantage of the esterification method is presented by side effects (partial hydrolysis of peptides, deamidation of aspar aginic and glutamic acid residues), which complicate the samples and the subsequent identification of pep tides. A comparison of the efficiency for various versions of IMAC can be found in papers [31, 32]. On an exam ple of protein extract NIH3T3, it was shown that, in the series Fe3+, Ga3+, Al3+, and Zr4+, gallium pos sessed the greatest selectivity toward phosphorylated proteins [33]. Generally, IMAC methods are utilized for the enrichment of phosphopeptides but can also help in the isolation of phosphorylated proteins [33] and, in addition, combine phosphoprotein and phos phopeptide isolation [34, 35]. Also, there are interest ing methods consisting in the sequential elution of phosphopeptides from IMAC and then allowing the separation of peptides by the number of phosphoryla tion sites [36]. In conclusion, it is worth saying that the IMAC methods possess advantages such as a relatively low cost, operational simplicity of the isolation process, compatibility with HPLC and mass spectrometry [37, 38], and opportunities to insert stable isotopes fol lowed by comparative analysis. At the same time, their basic drawback is represented by the unspecific bind ing of nonphosphorylated peptides, which compli cates the subsequent analysis. None of the modifica tions of the method aimed to increase its selectivity could fully eliminate unspecific binding. MOAC methods. The MOAC methods are based on the isolation of phosphopeptides on metal oxides (an inert supporter with the oxides of the respective metals applied onto its surface can also be used). Crystalline titanium oxide TiO2 finds the widest application [39– 41]. As a plausible mechanism for binding phospho peptides, researchers considered the interaction of two

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION

O

O P O

O

TiO2 TiO2 Fig. 4. Plausible mechanism for the binding of phospho peptides on the surface of crystalline TiO2.

titanium atoms on the surface of the solid phase with a phosphogroup of phosphopeptide (bridging mecha nism) (Fig. 4) [42]. In addition to TiO2, ZrO2 [43] and Al2O3 [44] were also used. Recently, the effective application of nio bium oxide Nb2O5 was reported [45]. It is worth men tioning that TiO2 and ZrO2 demonstrate different specificities toward phosphopeptides [46, 47]. The experiments with model proteins have shown that TiO2 is characterized by an enhanced affinity to pep tides bearing several phosphorylation sites, while ZrO2 basically retains monophosphorylated peptides [47]. Hence, the isolation of phosphopeptides on titanium and zirconium oxides can complement each other. Aluminum oxide can also be applied to the isolation of phosphoproteins. In work [44], on examples of model mixtures, the efficiency of the preconcentration of phosphoproteins and phosphopeptides on Al2O3 was demonstrated. Recently, we have prepared and tested a novel affine phase for the isolation of phosphopeptides using Sephadex G25 covalentlymodified by aluminum [48]. We used triethylaluminum for modification; it interacted with free hydroxyl groups on the surface of sephadex; in the subsequent hydrolysis, the retained alkyl groups were replaced by hydroxyl ones (Fig. 5).

Covalent immobilization ensures its strong binding on the phase, opening up an opportunity to vary the con ditions of isolation within wide ranges with no risk of metal elution together with phosphopeptides. The MOAC methods are strongly compatible with HPLC and mass spectrometry, which favors their use in phosphoproteomics. However, they also possess certain disadvantages. The main problem in the anal ogy to IMAC is created by the unspecific binding of acidic peptides. Nevertheless, a significant increase in the selectivity of phosphopeptide isolation upon the addition of 2,5dihydroxybenzoic acid to a mixture was reported in [42]. A similar effect was observed for phthalic [49] and glutamic [50] acids. These acids are assumed to competitively inhibit the interaction of peptide carboxyl groups with the phase. Cationexchange chromatography. The principle of cationexchange chromatography is based on the interaction of positively charged groups of peptides or proteins (amino groups of lysine, arginine, histidine, Nterminal amino groups) with the negatively charged surface of a stationary phase. This type of chromatog raphy is widely used in proteomics for the separation of proteins and peptides; moreover, it can be combined with reversedphase chromatography, making possible the separation of samples in two dimensions (Mud PIT means Multidimensional Protein Identification Technology) [51]. The use of this method in phosphoproteomics is based on the difference in the charge between phos phopeptides and nonphosphorylated peptides in acidic media (pH 2.7). The most nonphosphorylated peptides (obtained upon proteolysis using tripsin) under these conditions bear a charge of +2 and are preserved well in the cationexchanger phase, while the charge of phosphopeptides appears to be lower

O

O O H H HO H

1299

O H O OH

O H Et6Al2/hexane H O Et2Al H

O H O O

AlEt2

O H2O

O H H O (HO)2Al H

O H O O

Al(HO)2

Fig. 5. Modification of Sephadex G25 with triethylaluminum. At the first step, protons of hydroxyl groups are replaced with diethylaluminum residues. The subsequent hydrolysis leads to the replacement of the remaining ethyl groups with hydroxyl ones. JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

1300

KAN’SHIN et al.

O HO P OH X (N−term)R'

N H

SH X

OH–

R(C−term)

(N−term)R'

O

HS

N H

X

SH

R(C−term)

(N−term)R'

O

S R(C−term)

N H

O

Biotin tag Maleimide group responsible for modification

O

O

X (N−term)R'

N H

S

H N

N

S O

O O

O

O

R(C−term)

H N HN

S

N H Biotincontaining fragment

O Affine isolation of modified peptides and mass spectrometric analysis

Fig. 6. Isolation of phosphoproteins/phosphopeptides based on the βelimination of the phospho group followed by the addition of a nucleophile and a biotin group. In an alkaline medium, phospho groups are eliminated from residues of serine (X = H) and threonine (X = CH3), and the formed double bonds participate in the reaction of nucleophilic addition. The use of ethanedithiol allows the introduction of a thiol group into the peptide/protein composition; it is then modified by a maleinimidecontaining tag through the interaction with the SH group, and biotin, which ensures the isolation of the modified peptides by affine chro matography.

because of the presence of acidic phospho groups, which, in turn, leads to their lower retention in the phase [52]. Nonetheless, the method is characterized by poor specificity [53] and, as a rule, used as an addi tional step prior to the utilization of other methods of phosphopeptide enrichment, such as, for example, IMAC [22, 54] or MOAC [4, 55, 56]. Chemical modification of phospho groups. This series of methods utilizes chemical reactions specific for phospho groups in phosphoproteins and phospho peptides. Basically, the method aims to replace or modify phospho groups, which would allow the incor poration of a chemical group into the peptide/protein structure to assist its isolation and/or mass spectro metric identification. The main requirement con verges to the use of chemical reactions proceeding with high yields under mild conditions with no side processes. The methods based on the chemical modification of phospho groups can be subdivided into two groups, i.e., methods removing the phosphoric acid residue and those leaving it intact. The first group of methods is based on the βelimination of a phospho group fol lowed by the Michael addition of a nucleophile; the second group includes the formation of phosphoami dates. The βelimination of phosphate from phospho serine and phosphothreonine proceeds in an alkaline

medium and gives dehydroalanine and dehydrobu tanoic acid [57]. The resulting double carbon–carbon bond can be used to attach a group that would simplify the isolation of proteins premodified in this way. The most common approach involves biotin. In this approach, the modification is performed by ethanedithiol followed by the addition of biotin via the maleinimide group [58]. Finally, the modified pep tides can be isolated by affine chromatography on avi dine (Fig. 6). Note that prior to the application of the method, the mercapto groups of cysteine should be oxidized since the alkylated thiol groups are unstable under the conditions of βelimination. The direct iso lation of peptides bearing SH groups on solid phases with the immobilized iodoacetamide groups is also used. The common drawback of the methods based on the elimination of the phospho group appears to be the principal impossibility for analyzing phosphorylation on tyrosine. In addition, the glycosilated residues of serine and threonine are subjected to βelimination too; their treatment gives the same products as βelim ination from residues of phosphoserine and phospho threonine. In work [59], an approach was proposed based on the formation of phosphoamidates; it gives a chance to enrich phosphopeptides phosphorylated on serine, threonine, and tyrosine residues [60, 61]. The general

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION

1301

H2NC2H4OH/carbodiimide OPO3H2

OPO2HNHC2H4OH

COOH

tBocHN

CONHC2H4OH TFA

tBocHN tBoc OPO3H2 H2N

OPO3H2

COOH

tBocHN (H2NC2H4S)2/carbodiimide

CONHC2H4OH DTT

Mass spectrometry

OPO2HNHC2H4SH

OPO3H2 H2N

CONHC2H4OH

tBocHN

CONHC2H4OH Stationary phase

OPO2HNHC2H4S

TFA tBocHN

CONHC2H4OH

Isolation of modified peptides Fig. 7. Isolation of phosphopeptides by the phosphoamidate method. To avoid the reactions of inter and intramolecular conden sation, amino groups are blocked by tertbutoxycarbonyl protection (Boc). The subsequent treatment with ethanolamine in the presence of carbodiimide (condensation reagent) results in the formation of amide (carboxyl group) and phosphoamidates bonds (phospho group), the latter finally are hydrolyzed by trifluoroacetic acid (TFA). The sequential condensation with 1,6diamino 3,4dithiohexane in the presence of carbodiimide and reduction with dithiothreitol leads to the formation of thiol groups in each phosphorylated peptide/protein. These SH groups are used for the subsequent isolation on a stationary phase bearing iodoaceta mide groups. Then, phosphopeptides are recovered with a TFA solution.

scheme of modification is presented in Fig. 7. To pre vent the condensation reaction, the researchers first blocked amino (by tertbutoxycarbonyl protection) and carboxyl (by formation of amides with ethanola mine) groups of the peptides. Then, the phospho groups were converted into phosphoamidates with aminoethanethiol upon the action of carbodiimides. At the same time, the phospho group does not leave the peptide structure. The isolation of modified pep tides is carried out in a solid phase containing iodoac etate groups. Finally, the peptides are eluted from the phase with solutions of trifluoroacetic acid, which also recovers phospho groups. Among the evident disadvantages of the method, it is worth noting its high laborintensity (large number of steps) as well as the necessity of blocking chemically all amino and carboxyl groups to prevent inter and intramolecular condensation processes. A comparison of the efficiency of the existing methods for the enrichment of phosphopeptides and JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

phosphoproteins is a difficult task. The results of the analysis of model proteins mixtures, the phosphoryla tion of which is known in advance, cannot be extrapo lated to more complicated samples. At the same time, the comparative analysis of complex biological objects is troublesome since there is no fully described phos phoproteome that could serve as a standard. Even more, while performing such a comparison, one has to sort out the influence of factors that do not refer to the enrichment process (sample preparation, different conditions of chromatographic and mass spectromet ric analysis). In work [49], an attempt was made to compare the efficiency of IMAC methods (on the basis of Fe3+), phosphoamidate chemistry [61], and MOAC (based on TiO2 and the use of phthalic and 2,5dihydroxybenzoic acids to enhance selectivity). As samples, they used peptides (1.5 mg) obtained by the proteolysis of proteins from cells D. melanogaster Kc 167. The highest selectivity was shown by the IMAC and isolation via phosphoamidates, which allowed the researchers to identify more than No. 13

2010

1302

KAN’SHIN et al. Mass spectrometric analysis of phosphopeptides

Ionization:

Elimination of a phospho group:

ESI—negative mode; MALDI—alternative matrixes.

use of phosphatases; registration of a product ion; elimination of a neutral fragment.

Fragmentation: electron capture (ECD); electron transfer (ETD); multistep fragmentation CIDMSn.

Fig. 8. Mass spectrometry approaches most often used in the analysis of phosphoproteome.

500 phosphorylation sites, while the number of non phosphorylated peptides turned out to be 8 and 34, respectively. Note that the mutual coverage of results derived by different methods was approximately 35% only. Therefore, the methods are characterized by dif ferent selectivities toward different phosphopeptides and may effectively complement each other. Here, one may conclude that the optimal approach to get the fullest information on the phosphoproteome can be achieved by a combination of various methods for the enrichment of phosphoproteins and phosphopeptides followed by the combination of the results obtained. 3. MASS SPECTROMETRY IN THE ANALYSIS OF PROTEIN PHOSPHORYLATION Currently, mass spectrometry has become the main instrument in the analysis of phosphoproteome [5] (Fig. 5). This is favored by the high sensitivity, resolu tion, wide range of measurable concentrations, and high precision natural for mass spectrometry, as well as by the development of effective methods of the ioniza tion and fragmentation of proteins and peptides, and the possibility for combining mass spectrometric anal ysis with various separation techniques (for example, HPLC/MS) [62, 63]. Tandem mass spectrometry, which includes two stages of analysis at least, has found the widest appli cation in proteomics. At the first stage (MS), the mass to charge ratio (m/z) for the peptide ion is measured; at the second (MS2), the peptide ion is subjected to induced fragmentation, which is caused, for example, by collision with neutral atoms followed by the mea surement of m/z for the fragments formed. From the fragmentation spectrum obtained, one can extract information on the amino acid sequence of the pep tide and its phosphorylation. As a rule, this is done automatically by comparison with theoretical frag mentation spectra for peptides from the database of amino acid sequences of proteins [64] using software

packages, such as SEQUEST [65] and MASCOT [66]. If necessary, the ions obtained by the fragmentation of the ionized peptide can be subjected to additional enforced fragmentation (MSn). Because the predomi nant majority of experiments in phosphoproteomics include the analysis of peptides prepared by the pro teolysis of proteins, in this section of the review we restrict ourselves to the consideration of methods of phosphoproteome analysis at the peptide level. Ionization of phosphopeptides. Since mass spec trometry determines ions in the gas phase, an impor tant role in proteomics is played by methods of protein and peptide ionization. Today, two groups of methods are known, electrospray ionization (ESI) and matrix assisted laser desorption/ionization (MALDI). Both methods can be utilized for the analysis of phosphop roteome; however, the determination of phosphoryla tion sites is easier for doubly and triply charged ions formed in ESI (MALDI basically results in the forma tion of singly charged ions). Moreover, ESI can be directly connected to HPLC. The general problem in the analysis of phospho peptides in the mode of positive ion registration is rep resented by the effects of the suppression of ionization in the presence of nonphosphorylated peptides. The preliminary enrichment of phosphopeptides allows one to solve the problem. Also one can use the mode of the registration of negative ions, but therein the fragmentation of phosphopeptides strongly deterio rates. In some cases, the effects of ionization suppres sion in MALDI can be diminished by using a mixture of 2,4,6trihydroxyacetophenone with diammonium citrate as a matrix [67]. For instance, it was shown that the ionization degree of phosphopeptides in such a matrix is approximately 10fold higher in comparison to the traditional matrixes, like αcyano3(4hyd roxyphenyl)acrylic acid. Here, the effect is enhanced with an increase in the number of phosphorylation sites in the peptide. The matrix has no significant influence on the character of phosphopeptide frag mentation. There is also an interesting approach exploiting matrices with immobilized metals to pre concentrate phosphopeptides [68]. Alternative methods for the fragmentation of phos phopeptides. As mentioned above, the main source of structural information on peptides is provided by MS2 spectra obtained by the fragmentation of ionized pep tides. In most cases, fragmentation is realized by the collision of peptide ions with uncharged atoms of an inert gas (helium or argon) inside the mass spectro meter (CID). Therein, the peptide ion decomposes via amide bonds into fragments to give yrange ions as basic products (the charge is located at the C end) and brange ions (the charge is located at the N end) (Fig. 9). Unfortunately, this method of fragmentation appears inefficient for phosphopeptides. Under CID conditions, the usual path of phosphopeptide frag mentation is due to the elimination of a phosphate

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION x3 y3

z3

x2 y2

z2

x1 y1

z1

and the radical cation formed undergoes subsequent fragmentation (Fig. 10). Today, the method is compa tible to instruments on the platform of ioncyclotron resonance with Fourier transformation (FTICR) exclusively, which are characterized by extremely pre cise mass registration [75]. In the ETD method, radi cals formed upon the chemical ionization of methane are used for electron transfer to peptide ions. ETD can be adopted for a wider range of instruments. In work [76], the IMACisolation of phosphopeptides from a sample of S. cerevisiae was combined with mass spec trometric analysis on an installation including a two dimensional ion trap with Fourier transformation (2DIT FT, 2Dimensional Ion Trap with Fourier Transform) and ETD fragmentation. Therein, using 30 μg of proteins only, the study succeeded in the iden tification of more than 1200 phosphorylation sites corresponding to 629 phosphoproteins. In the other investigation [77], phosphopeptides isolated by MOAC on the basis of TiO2 were analyzed on an instrument including a threedimensional ion trap (3DIT, 3Dimensional Ion Trap) with ETD fragmen tation; this resulted in the discovery of 1435 phospho rylation sites, 80% of which were unknown before. One of the ETD modifications utilizes an additional activation method (collisions with neutral atoms), which allows the more effective fragmentation of pep tides with the charge +2 [78]. Also, the efficient use of an ETD and ECD combination was reported [79].

H+

R2 O R4 O R1 O R3 O H2N CH C HN CH C HN CH C HN CH C OH a1 b1

c1

a2 b2

c2

a3 b3

c3

Fig. 9. Nomenclature of ions formed upon the peptides fragmentation.

group, which is accompanied by some minor fragmen tation via the carbon skeleton. Thus, the MS2 spec trum obtained contains little structural information and cannot be used for the identification of the amino acid sequence of peptides. Alternative fragmentation methods were developed in [69]; these are especially convenient for the deter mination of posttranslational modifications (phos phorylation, glycosidation, acylation, and sulfona tion) unstable under CID [70]. These include frag mentation due to electron capture (ECD) and electron transfer (ETD) [71]. The core of both me thods converges to the capture/transfer of an electron to a peptide ion, which then undergoes fragmentation differing from that in CID and leading to c and z ions (Fig. 10) [72–74]. At the same time, the phospho group is not eliminated and there appears an opportu nity to determine not only the amino acid sequence of the peptide but also its phosphorylation sites. In the ECD method, a lowenergy electron is transferred directly to a positively charged peptide ion,

The necessity of expensive equipment, which is not affordable for many mass spectrometry laboratories, should be included among the disadvantages of the ECD and ETD methods. •

+

H2N e– + H

N H

…

O

R +

N H

NH3

O

H N

OH

+

O R

H N NH

+

NH 3

+

O

H

NH3

N H

H N

O OH

O

+

NH3

H2N H

O

R

OH

+

O •

+

O

c

O

R

NH2

NH3

N H

H N

NH3

O OH

O

z +

+

NH3

NH3

Fig. 10. Capture of an electron by a peptide ion and its subsequent fragmentation. JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

1303

No. 13

2010

1304

KAN’SHIN et al.

Use of multistep fragmentation MSn (n > 2). As mentioned before, often phosphopeptides reveal poor informative fragmentation under CID conditions. The major fragmentation path in MS2 appears to be the elimination of the phosphate group. However, there is a chance of the consequent isolation and analysis of ions formed in the first fragmentation, i.e., MS3. This provides an opportunity to obtain information about the peptide structure even without satisfactory frag mentation at the first step [52, 55, 80]. Experiments with multistep fragmentation can be performed on widely used instruments with iontrap analyzers. A socalled pseudo MSnapproach has also been proposed. Herein, the ions are not separated before the next fragmentation step. Instead, the fragment formed by the elimination of a phosphate group undergoes additional activation and CID fragmenta tion, and the sum spectrum thus recorded is used for the determination of the peptide structure [55, 81]. Methods for the registration of product ions and neutral fragments. The fragmentation of phosphopep tides (registration of negative ions) under the condi tions of CID involves the formation of the characteris – tic PO 3 ion with m/z 79. This ion can be used as a marker of the phospho group, which allows the identi fication of phosphorylated peptides. The amino acid sequence of peptides can be established later using the fragmentation of their positive ions [34]. Another opportunity consists in switching between the modes of the registration of positive and negative ions within the same experiment. Thus, if an ion with m/z 79 is detected, the system is automatically switched to the positive ion mode and the fragmentation of the respec tive peptide is registered [34, 82, 83]. The peptides containing phosphotyrosine produce the characteristic immonium ion with m/z 216.043 upon fragmentation; it corresponds to the elimination of phosphotyrosine. Thus, the detection of this ion in the spectrum can be used for the selective isolation of peptides phosphorylated on tyrosine. In work [84], the method made possible the highly sensitive determina tion of phosphopeptides at the picomolar level; the proteins were separated by gel electrophoresis. The scanning of precursor ions appears to be an effective approach to the analysis of phosphorylation in complex mixtures containing both phosphorylated and nonphosphorylated peptides, i.e., with no pre liminary enrichment of phosphopeptides. The method of neutral fragment elimination (neu tral loss scanning) differs slightly from the method of precursor scanning and includes the detection of cer tain mass losses by peptides. It is based on the determi nation of phosphoric acid by the elimination of ortho phosphoric acid (H3PO4, m/z 98) for phosphoserine and phosphothreonine and methaphosphoric acid (HPO3, m/z 80) for phosphotyrosine under CID con ditions with the registration of positive ions.

The methods described are frequently performed on instruments with a triple quadrupole analyzer. Nonetheless, one has to take into account the errors arising because of the complexity of the MS2 spectrum and those caused by restrictions in the precision of mass determination and the resolution of these instru ments. A considerable improvement was made by the incorporation of hybrid quadrupole timeofflight analyzers (QTOF), which are characterized by a higher resolution [84]. Use of enzymatic dephosphorylation. The method is based on the comparative mass spectrometric analysis of the change of the peptide mass upon the action of phosphatases, enzymes capable of efficiently dephos phorylating proteins and peptides. The phosphatase treatment of samples removes phospho groups from the proteins and peptides and reduces their mass by 80n Da, where n means the number of phospho groups. The mass spectrometric analysis of the same sample before and after enzymatic dephosphorylation helps to reveal ions that underwent shifts along the m/z scale, which provides a tool for the determination of which peptides were phosphorylated and how many phospho groups they contain [85]. The method under consideration seems analogous to the method described above of neutral fragment loss with the dif ference that the phosphate elimination proceeds under the action of phosphatases. However, there is a disadvantage that the ionization of phosphopeptides is suppressed in the presence of nonphosphorylated peptides, which complicates the identification of sig nals corresponding to phosphopeptides in a non dephosphorylated sample. 4. COMPARATIVE PHOSPHOPROTEOMICS Along with the determination of phosphoproteins and the sites of their phosphorylation, an important task is represented by the comparison of phosphory lated proteins in different samples. The comparison can be used, for example, for the quantitative evalua tion of changes in the phosphorylation of proteins in cells upon the action of various outer effects (hor mones, medications, changes of temperature, etc.) or in biological processes. In most papers, the compara tive analysis of phosphopeptide concentrations among the samples was performed. However, an important task is also provided by the comparison of phosphory lated and nonphosphorylated species of the very same peptide, i.e., the stoichometry of phosphorylation. Despite the obvious value of this information, it has not attracted sufficient attention so far, because the methods used for the isolation of phosphopeptides and phosphoproteins cannot be used here [86–88]. The methods allowing the comparison of phos phoproteomes can be subdivided into two groups: semiquantitative methods without using isotope tags, and quantitative ones including such tagging. It is especially worth mentioning a method based on the

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION

1305

Heavy Protein A in Isotope sample 1 insertion

Proteolysis

Isolation on matrix Peptides mixture

Light Protein A in sample 2 Rinse

Desorption

Mass spectrometry

m/z

Fig. 11. Principles of isotope labeling.

use of internal standards, which allows the precise measurement of peptide and protein concentrations in the sample [89]. The detailed description of different approaches in comparative proteomics can be found in [90–96]. In this review, we will pay attention only to those applica ble in the analysis of phosphoproteome. Semiquantitative methods. The semiquantitative methods converge to the consequent analysis of a number of samples followed by the comparison of the results obtained. The comparison utilizes the intensi ties of peptide signals in the mass spectrum; at the same time, the intensity of a signal is considered to be directly proportional to the peptide concentration. Software packages were elaborated to accomplish such analysis; these include processing steps, such as the isolation of a peak corresponding to a peptide, the removal of noise signals, and the comparison of chro matograms obtained for different samples [97, 98]. However, the intensity of signals is also affected by other factors, and the main goal of such experiments is to minimize their influence. Unfortunately, it is really difficult to reproduce exactly the same conditions in mass spectrometric experiments; even in the case when the samples are analyzed on the very same instrument with identical parameters, the ionization of peptides may differ. Also, it is crucial to extinguish all differences between the steps performed before mass spectrometry, i.e., the extraction of proteins, proteolysis, the enrichment of phosphopeptides, and chromatography. Papers [35, 99] can be mentioned as application examples of semiquantitative methods in proteomics. Another way to compare concentrations is based on the correlation between the quantity of the peptides detected and the concentration of the respective pro tein (PAIs, Protein Abundant Indices) [100]. The relation was proved to be logarithmic [101]. In conclu sion, one might say that the semiquantitative approach JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

seems really simple but is characterized by rather low precision. Quantitative methods. As mentioned above, a com parison based on the sequential analysis of several samples is not very precise. The perfect variant would have been the simultaneous analysis of samples, which could avoid completely all side interferences on the ratio of signal intensities. However, such analysis requires a certain modification of peptides, which would then allow the analysis to distinguish signals corresponding to the same peptides from different samples. For instance, there is a chance to change masses of peptides referring to different samples (Fig. 11). At the same time, the comparison of inten sities for light and heavy forms of the very same pep tide is used for comparative analysis. Note that the peptide modification should create a difference in masses exclusively, leaving all other parameters intact, such as ionization efficiency, charge, chromatographic behavior, and character of fragmentation. All these requirements are fulfilled for a modification that uti lizes various stable isotopes; the most widely used are couples 1H/2H(D), 12C/13C, 14N/15N, and 16O/18O. There are two common methods of isotope inser tion into proteins and peptides: in vivo insertion of iso topes generally using various isotope forms of amino acids and the chemical modification of proteins and peptides by substances bearing various isotopes. In vivo isotope modification of proteins. The SILAC method, which is based on the cultivation of cells in a nutritional medium in the presence of various isotopic forms of amino acids, which are then incorporated into the peptide due to natural metabolism, was elab orated in 1999 [102]; it is now widely used for the quantitative comparison of proteomes and phosphop roteomes [103, 104]. This method most often utilizes isotopically tagged lysine and arginine, because in view of the sub strate specificity of tripsin, proteolysis gives peptides No. 13

2010

1306

KAN’SHIN et al.

containing at least a singular residue of premodified amino acids. Cells are usually grown in several cycles, which leads to the virtually complete incorporation of the marked amino acids into proteins. Then, the cells are crushed, the proteins are extracted, and different samples are combined before the subsequent analysis, which includes proteolysis, separation, the isolation of phosphopeptides, and mass spectrometry. The mass spectrometric analysis is capable of measuring the ratio of light and heavy forms of the peptide. A combination of SILAC with various techniques of phosphoprotein and phosphopeptide isolation is also widely used. For instance, the method was suc cessfully combined with the immunoprecipitation of proteins [105–107] and phosphopeptides [108] for the analysis of phosphorylation on tyrosine. In paper [14], the analysis of serine and threonine phosphorylation was performed using a combination of SILAC, cation exchange chromatography, and IMAC. Recently, a detailed analysis of the influence of the epidermal growth factor on the protein phosphorylation in HeLa cells by means of SILAC [55] has been published; the changes in phosphorylation were registered as a func tion of time (five time intervals) for cytosol and nucleus peptide fractions. Therein, to isolate phos phopeptides, a combination of cationexchange chro matography and MOAC on basis of TiO2 was used, while the mass spectrometric analysis was performed on an apparatus consisting of a twodimensional qua drupole with Fourier transformation. As a result, 6600 phosphorylation sites were identified; for 14% of these, phosphorylation changed upon the action of the growth factor. Investigations using SILAC are generally restricted to the use of cell cultures; however, experiments with flies, worms [109], and rats [110] were also reported. One needs to respect the high cost of isotopetagged amino acids, especially in experiments on the modifi cation of whole organisms. Chemical modification of proteins and peptides by isotope tags. These approaches are based on the chem ical modification of proteins and peptides by the reagents available in several isotopic forms, the so called isotope tags. The first isotope affine tags (ICAT) were developed in 1999 [111]. In the structure of the tags, one can dis tinguish three fragments, a group reacting with cys teine selectively, isotopecontaining groups with eight hydrogen or deuterium atoms, and a biotin fragment facilitating the subsequent isolation. The proteins are modified on cysteine residues by light and heavy forms of ICAT and subjected to proteolysis; then, the modi fied peptides can be isolated by affine chromatography on avidine. This method makes the peptide mixture simpler, since only those peptides that contain cys teine are subjected to analysis; however, there is no chance for the quantitative analysis of proteins with no cysteine in their composition. In 2003, socalled

cleavable ICAT (cICAT) was elaborated [112], in which the 12C/13C isotopes were used to decrease the spread between the retention times in a reversed phase among the isotopic forms, in contrast to versions based on 1H/2H [113]. Also cICAT provided an opportunity to remove the biotin fragment in an acidic medium prior to the peptide analysis. The methods based on ICAT are not widely applied for the quantitative study of phosphopeptides, as they are limited to the analysis of cysteinecontaining pep tides, while the chance for the presence of both the phospho group and cystein in peptide at the same time is small. However, ICAT has been applied for the com parative quantitative analysis of phosphoproteins in combination with their selective isolation (for exam ple, by immunoprecipitation [114]). Another method of the chemical modification of peptides applies isobaric tags (iTRAQ) [115]. The tags react selectively with amino groups of peptides, there fore assuring the full coverage of the proteome. At the same time, the comparative analysis is implemented at the fragmentation step (in MS2). The particular fea ture of the tags is that they have the same mass and thus disable the recognition of the modified peptide isoto pic forms in the MS spectrum. However, despite the identical mass, they differ by their isotopic composi tion (appear as moleculesisobars), which leads to the appearance of different ions in the fragmentation spectrum of the modified peptide and allows quantita tive determination. In the fragmentation, the tag iTRAQ does not dissociate or affect the subsequent fragmentation of the peptide. The cleavage of the tag itself gives signal ions (reporter ions) differing in mass for different forms of the tag. The comparison of the intensities for the signal ions in MS2 is utilized for quantitative analysis. The signal ions are registered in the region of low masses (from 113 to 121) and do not overlap with the ions derived from peptide fragmenta tion. For the time being, iTRAQ tags have been deve loped that allow comparison among eight samples. Unfortunately, the iTRAQ analysis cannot be per formed using instruments with iontrap analyzers because of the respective restrictions for measuring low m/z values. As an example of using iTRAQ for the quantitative analysis of phosphopeptides, we note paper [116], the authors of which investigated phosphorylation in mouse synaptosomes upon the action of sodium chlo ride. For the isolation of phosphopeptides, they used a combination of cationexchange chromatography and IMAC. Tyrosine phosphorylation is studied widely using a combination of iTRAQ with immunoprecipi tation [117–119]. There are also strategies for the selective insertion of isotopic tags into phosphoproteins based on the βelimination of the phospho group followed by the Michael addition of nucleophiles. PhIAT tags [120] and their modifications, which use isolation on the

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION

PhIST stationary phase [121], refer to such reagents; however, they have not been widely accepted because of the limitations described above for the β−elimina tion reaction. CONCLUSIONS The determination of phosphorylated proteins, sites of their phosphorylation, the stoichometry of phosphorylation for those sites, and the quantitative comparison of phosphoproteomes appear extremely actual tasks today. Currently, many effective approaches have been developed, which allow the exceptional analysis of phosphorylation by mass spectrometry. At the same time, one of the basic directions appears to be the selective isolation of phosphoproteins and phospho peptides; this allows one to concentrate on the analysis on phosphoproteome. Also, attention is paid to the reduction of the number of analytical steps, which would help to decrease sample losses and enhance sen sitivity. Also, new mass spectrometric approaches were proposed to simplify the analysis of phosphopeptides and phosphoproteins. Among them, we should first mention alternative methods of fragmentation, such as ETD and ECD. For the time being, phosphoproteomic experi ments provide us with a huge amount of information, which cannot be processed manually. Therefore, vari ous software packages aimed were elaborated to assist the analysis of mass spectrometric data and provide information on the phosphopeptides present in sam ples and related phosphoproteins [55, 122]. Also, there exist PC programs for the subsequent annotation for phosphoproteins. Vast databases containing infor mation on protein phosphorylation became available, for example PhosphoSite [123], now listing around 50000 phosphorylation sites. To make the proteome data accessible for the scientific community, various depositories that contain not only the processed data only but also the initial mass spectrometry acquisition files were designed [124, 125]. Meanwhile, all the developed methods have their own restrictions in specificity, sensitivity, etc.; there fore, in our opinion, the main direction for the evolu tion of isolation tools for phosphoproteins and phos phopeptides will be represented by the combination of several different methods for the examination of the same sample followed by the combination of the results obtained. As it was shown [18], such an approach provides the detection of different partially overlapped regions of phosphoproteome, giving an opportunity to look at it from several points of view. A similar situation is formed with various mass spectro metric approaches; the combination of data acquired on different instruments by different methods of ion ization (ESI, MALDI) and fragmentation (CID, ETD, ECD, MS3) helps to substantially increase the JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

1307

number of peptides being charted. The wider accep tance will also be gained by the methods of the two step isolation of phosphoproteins and phosphopep tides, which would allow the analysis of not only phos phopeptides but also of nonphosphorylated peptides corresponding to the phosphorylated ones; this would ensure the more reliable identification and compara tive quantitative analysis. In addition to the detection of the expanding quantity of phosphorylated proteins and an increase in sensitivity, the progress in phos phoproteomics will correspond to the determination of the stoichometry for protein phosphorylation and its alteration upon the action of various outer factors or in biological processes. Under the contemporary conditions, when phos phoproteomic studies are performed in many labora tories, and an enormous mass of data is generated, the question of their standardization is becoming crucially important. The use of mass spectrometers from differ ent manufacturers as well as different software pack ages for the processing of the data obtained creates problems in the comparison of the results acquired in different groups [126, 127]. Nowadays, general for mats for the representation of mass spectral data and the standardization of their analysis are being actively developed [128]. REFERENCES 1. Kalume, D.E., Molina, H., and Pandey, A., Curr. Opin. Chem. Biol., 2003, vol. 7, no. 1, p. 64. 2. Johnson, S.A. and Hunter, T., Nat. Methods, 2005, vol. 2, no. 1, p. 17. 3. Sickmann, A. and Meyer, H.E., Proteomics, 2001, vol. 1, no. 2, p. 200. 4. Olsen, J.V., Blagoev, B., Gnad, F., Macek, B., et al., Cell, 2006, vol. 127, no. 3, p. 635. 5. Zhou, M. and Veenstra, T., Biotechniques, 2008, vol. 44, no. 4, p. 670. 6. Ham, B.M., Yang, F., Jayachandran, H., Jaitly, N., et al., J. Proteome Res., 2008, vol. 7, no. 6, p. 2215. 7. Issaq, H. and Veenstra, T., Biotechniques, 2008, vol. 44, no. 5, p. 697. 8. Van den Bergh, G. and Arckens, L., Methods Mol. Biol., 2008, vol. 424, p. 147. 9. Huck, C.W. and Bonn, G.K., Methods Mol. Biol., 2008, vol. 384, p. 507. 10. Dolník, V., Electrophoresis, 2008, vol. 29, no. 1, p. 143. 11. Hunter, T., Cell, 1995, vol. 83, no. 1, p. 1. 12. Schlessinger, J., The Harvey Lectures, Ser.: 89, 1993, p. 105. 13. Hemmings, Jr.H.C., Neuromethods, 1996, vol. 30, p. 121. 14. Gruhler, A., Olsen, J.V., Mohammed, S., Martensen, P., et al., Mol. Cell Proteomics, 2005, vol. 4, no. 3, p. 310. 15. Mann, M., Ong, S.E., Gronborg, M., Steen, H., et al., Trends Biotechnol., 2002, vol. 20, no. 6, p. 261. 16. Zhang, H., Zha, X., Tan, Y., Hornbeck, P.V., et al., J. Biol. Chem., 2002, vol. 277, no. 42, p. 39379. No. 13

2010

1308

KAN’SHIN et al.

17. Hattori, S., Lida, N., and Kosako, H., Expert Rev. Proteo mics, 2008, vol. 5, no. 3, p. 497. 18. Gronborg, M., Kristiansen, T.Z., Stensballe, A., Andersen, J.S., et al., Mol. Cell Proteomics, 2002, vol. 1, no. 7, p. 517. 19. Steen, H., Kuster, B., Fernandez, M., Pandey, A., and Mann, M., J. Biol. Chem., 2002, vol. 277, no. 2, p. 1031. 20. Bergström, L.S., Molin, M., Savitski, M.M., Emilsson, L., et al., J. Proteome Res., 2008, vol. 7, no. 7, p. 2897. 21. Sugiyama, N., Nakagami, H., Mochida, K., Daudi, A., Tomita, M., Shirasu, K., and Ishihama, Y., Mol. Syst. Biol., 2008, vol. 4, p. 193. 22. Villen, J., Beausoleil, S.A., Gerber, S.A., and Gygi, S.P., Proc. Natl. Acad. Sci. USA, 2007, vol. 104, no. 5, p. 1488. 23. Ficarro, S., Chertihin, O., Westbrook, V.A., White, F., et al., J. Biol. Chem., 2003, vol. 278, no. 13, p. 11579. 24. Porath, J., Carlsson, J., Olsson, I., and Belfrage, G., Nature, 1975, vol. 258, no. 5536, p. 598. 25. Michel, H.P. and Bennett, J., J. Biol. Chem., 1987, vol. 263, no. 3, p. 1123. 26. Li, Y., Lin, H., Deng, C., Yang, P., and Zhang, X., Proteo mics, 2008, vol. 8, no. 2, p. 238. 27. Liu, H.L., Ho, Y., and Hsu, C.M., J. Biomol. Struct. Dyn., 2003, vol. 21, no. 1, p. 31. 28. Nuhse, T.S., Stensballe, A., Jensen, O.N., and Peck, S.C., Mol. Cell. Proteomics, 2003, vol. 2, no. 11, p. 1234. 29. Posewitz, M.C. and Tempst, P., Anal. Chem., 1999, vol. 71, no. 14, p. 2883. 30. Ficarro, S.B., McCleland, M.L., Stukenberg, P.T., Burke, D.J., et al., Nat. Biotechnol, 2002, vol. 20, no. 3, p. 301. 31. Kånge, R., Selditz, U., Granberg, M., Lindberg, U., Ekstrand, G., and Gustafsson, M., J. Biomol. Tech., 2005, vol. 16, no. 2, p. 91. 32. Machida, M., Kosako, H., Shirakabe, K., Kobayashi, M., Ushiyama, M., Inagawa, J., Hirano, J., Nakano, T., Bando, Y., Nishida, E., and Hattori, S., FEBS J., 2007, vol. 274, no. 6, p. 1576. 33. Collins, M.O., Yu, L., Husi, H., Blackstock, W.P., et al., Sci. STKE, 2005, vol. 298, p. 6. 34. Collins, M.O., Yu, L., Coba, M.P., Husi, H., et al., J. Biol. Chem., 2005, vol. 280, no. 7, p. 5972. 35. Gutillas, P.R., Geering, B., Waterfield, M.D., and Van haesebroeck, B., Mol. Cell Proteomics, 2005, vol. 4, no. 8, p.1038. 36. Thingholm, T.E., Jensen, O.N., Robinson, P.J., and Larsen, M.R., Mol. Cell Proteomics, 2008, vol. 7, no. 4, pp. 661. 37. Riggs, L., Sioma, C., and Regnier, F.E., J. Chromatogr. A, 2001, vol. 924, nos. 1–2, p. 359. 38. Raska, C.S., Parker, C.E., Dominski, Z., and Marzluff, W.F., Anal. Chem., 2002, vol. 74, no. 14, p. 3429. 39. Pinkse, M.W., Uitto, P.M., Hilhorst, M.J., Ooms, B., and Heck, A.J., Anal. Chem., 2004, vol. 76, no. 14, p. 3935. 40. Li, Y., Xu, X., Qi, D., Deng, C., Yang, P., and Zhang, X., J. Proteome Res., 2008, vol. 7, no. 6, p. 2526.

41. Mohammed, S., Kraiczek, K., Pinkse, M.W.H., Lemeer, S., Benschop, J.J., and Heck, A.J.R., J. Proteome Res., 2008, no. 4, p. 1565. 42. Larsen, M.R., Thingholm, T.E., Jensen, O.N., Roepstorff, P., and Jorgensen, T.J., Mol. Cell Proteomics, 2005, vol. 4, no. 7, p. 873. 43. Kweon, H.K. and Hakansson, K., Anal. Chem., 2006, vol. 78, no. 6, p. 1743. 44. Wolschin, F., Wienkoop, S., and Weckwerth, W., Proteomics, 2005, vol. 5, no. 17, p. 4389. 45. Ficarro, S.B., Parikh, J.R., Blank, N.C., and Marto, J.A., Anal. Chem., 2008, vol. 80, no. 12, p. 4606. 46. Jensen, S.S. and Larsen, M.R., Rapid Commun. Mass Spec trom., 2007, vol. 21, p. 3635. 47. Kweon, H.K. and Håkansson, K., Anal. Chem., 2006, vol. 78, no. 6, p. 1743. 48. Kan’shin, E.D., Nifant’ev, I.E., and Pshezhetskii, A.V., Massspektrometria, 2008, vol. 5, no. 3, p. 195. 49. Bodenmiller, B., Mueller, L.N., Mueller, M., Domon, B., and Aebersold, R., Nat. Methods, 2007, vol. 4, no. 3, p. 231. 50. Yu, L.R., Zhu, Z., Chan, K.C., Issaq, H.J., Dimitrov, D.S., and Veenstra, T.D., J. Proteome Res., 2007, vol. 6, no. 11, p. 4150. 51. Link, A.J., Fng, J., Schieltz, D.M., Carmack, E., et al., Nat. Biotechnol, 1999, vol. 17, no. 7, p. 676. 52. Beausoleil, S.A., Jedrychowski, M., Schwartz, D., Elias, J.E., et al., Proc. Natl. Acad. Sci. USA, 2004, vol. 101, no. 33, p. 12130. 53. Trinidad, J.C., Specht, C.G., Thalhammer, A., Schoep fer, R., and Burlingame, A.L., Mol. Cell Proteomics, 2006, vol. 5, no. 5, p. 914. 54. Zhai, B., Villén, J., Beausoleil, S.A., Mintseris, J., and Gygi, S.P., J. Proteome Res., 2008, vol. 7, no. 4, p. 1675. 55. Olsen, J.V., Blagoev, B., Gnad, F., Macek, B., et al., Cell, 2008, vol. 127, no. 3, p. 635. 56. Sui, S., Wang, J., Yang, B., Song, L., Zhang, J., Chen, M., Liu, J., Lu, Z., Cai, Y., Chen, S., Bi, W., Zhu, Y., He, F., and Quan, X., Proteomics, 2008, vol. 8, no. 10, p. 2024. 57. Jaffe, H. and Veeranna, H.C., Biochemistry, 1998, vol. 37, no. 46, p. 16211. 58. Oda, Y., Nagasu, T., and Chait, B.T., Nat. Biotechnol, 2001, vol. 19, no. 4, p. 379. 59. Zhou, H., Watts, D., and Aebersold, R., A Nat. Biotechnol., 2001, vol. 19, no. 4, p. 375. 60. Tao, W.A., Wollscheid, B., O’Brien, R., Eng, J.K., et al., Nat. Methods, 2005, vol. 2, no. 8, p. 591. 61. Bodenmiller, B., Mueller, L.N., Pedrioli, P.G., Pflieger, D., et al., Mol. Biosyst., 2007, vol. 3, no. 4, p. 275. 62. Domon, B. and Aebersold, R., Science, 2006, vol. 312, no. 5771, p. 212. 63. Loyet, K.M., Stults, J.T., and Arnott, D., Mol. Cell Proteo mics, 2003, vol. 4, no. 3, p. 235. 64. McHugh, L. and Arthur, J.W., PLoS Comput. Biol., 2008, vol. 4, no. 2, p. 12. 65. Ducret, A., Van Oostveen, I., Eng, J.K., Yates, J.R., and Aebersold, R., Protein Sci., 1998, vol. 7, no. 3, p. 706.

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010

MASS SPECTROMETRIC ANALYSIS OF PROTEIN PHOSPHORYLATION 66. Perkins, D.N., Pappin, D.J., Creasy, D.M., and Cottrell, J.S., Electrophoresis, 1999, vol. 20, no. 18, p. 3551. 67. Yang, X., Wu, H., Kobayashi, T., Solaro, R.J., and van Bree men, R.B., Anal. Chem., 2004, vol. 76, no. 5, p. 1532. 68. Ibanez, A.J., Muck, A., and Svatos, A., J. Proteome Res., 2007, vol. 6, no. 9, p. 3842. 69. Zubarev, R.A., Zubarev, A.R., and Savitski, M.M., J. Am. Soc. Mass Spectrom., 2008, vol. 19, no. 6, p. 753. 70. Cooper, H.J., Hakansson, K., and Marshall, A.G., Mass Spectrom. Rev., 2005, vol. 24, no. 2, p. 201. 71. Syka, J.E., Coon, J.J., Schroeder, M.J., Shabanowitz, J., and Hunt, D.F., Proc. Natl. Acad. Sci. USA, 2004, vol. 101, no. 26, p. 9528. 72. Zubarev, R.A., Curr. Opin Biotechnol., 2004, vol. 15, no. 1, p. 12. 73. Stensballe, A., Jensen, O.N., Olsen, J.V., Haselmann, K.F., and Zubarev, R.A., Rapid Commun. Mass Spectrom., 2000, vol. 14, no. 19, p. 1793. 74. Rand, K.D., Adams, C.M., Zubarev, R.A., and Jør gensen, T.J., J. Am. Chem. Soc., 2008, vol. 130, no. 4, p. 1341. 75. Sweet, S.M., Creese, A.J., and Cooper, H.J., Anal. Chem., 2006, vol. 78, no. 21, p. 7563. 76. Chi, A., Huttenhower, C., Geer, L.Y., Coon, J.J., et al., Proc. Natl. Acad. Sci. USA, 2007, vol. 104, no. 7, p. 2193. 77. Molina, H., Horn, D.M., Tang, N., Mathivanan, S., and Pandey, A., Proc. Natl. Acad. Sci. USA, 2007, vol. 104, no. 7, p. 2199. 78. Swaney, D.L., McAlister, G.C., Wirtala, M., Schwartz, J.C., et al., Anal. Chem., 2007, vol. 79, no. 2, p. 477. 79. Wu, S.L., Huhmer, A.F.R., Hao, Z., and Karger, B.L., J. Proteome Res., 2008, vol. 6, no. 11, p. 4230. 80. Olsen, J.V. and Mann, M., Proc. Natl. Acad. Sci. USA, 2004, vol. 101, no. 37, p. 13417. 81. Schroeder, M.J., Shabanowitz, J., Schwartz, J.C., Hunt, D.F., and Coon, J.J., Anal. Chem., 2004, vol. 76, no. 13, p. 3590. 82. Zappacosta, F., Huddleston, M.J., Karcher, R.L., Gelfand, V.I., et al., Anal. Chem., 2002, vol. 74, no. 13, p. 3221. 83. Williamson, B.L., Marchese, J., and Morrice, N.A., Mol. Cell Proteomics, 2006, vol. 5, no. 2, p. 337. 84. Steen, H., Kuster, B., and Mann, M., J. Mass Spectrom., 2001, vol. 36, no. 7, p. 782. 85. Larsen, M.R., Seornsen, G.L., Fey, S.J., and Larsen, P.M., Proteomics, 2001, vol. 1, no. 2, p. 223. 86. Steen, H., Jebanathirajah, J.A., Springer, M., and Kir schner, M.W., Proc. Natl. Acad. Sci. USA, 2005, vol. 120, no. 11, p. 3948. 87. Bonenfant, D., Schmelzle, T., Jacinto, E., Crespo, J.L., Mini, T., Hall, M.N., and Jenoe, P., Proc. Natl. Acad. Sci. USA, 2003, vol. 100, no. 3, p. 880. 88. Hegeman, A.D., Harms, A.C., Sussman, M.R., Bunner, A.E., and Harper, J.F., J. Am. Soc. Mass Spectrom., 2003, vol. 15, no. 5, p. 647. JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

1309

89. Gerber, S.A., Rush, J., Stemman, O., Kirschner, M.W., and Gygi, S.P., Proc. Natl. Acad. Sci. USA, 2003, vol. 100, no. 12, p. 6940. 90. Putz, S., Reinders, J., Reinders, Y., and Sickmann, A., Expert Rev. Proteomics, 2005, vol. 2, no. 3, p. 381. 91. Ong, S.E. and Mann, M., Nat. Chem. Biol, 2005, vol. 1, no. 5, p. 252. 92. Ding, S.J., Qian, W.J., and Smith, R.D., Expert Rev. Pro teomics, 2007, vol. 4, no. 1, p. 13. 93. Aggarwal, K., Choe, L.H., and Lee, K.H., Brief Funct. Genomic Proteomic, 2006, vol. 5, no. 2, p. 112. 94. Unwin, R.D., Evans, C.A., and Whetton, A.D., Trends Bio chem. Sci., 2006, vol. 31, no. 8, p. 473. 95. Chen, X., Sun, L., Yu, Y., Xue, Y., and Yang, P., Expert Rev. Proteomics, 2007, vol. 4, no. 1, p. 25. 96. Mann, M., Nat/ Rev. Mol. Cell Biol, 2006, vol. 7, no. 12, p. 952. 97. Ono, M., Shitashige, M., Honda, K., Isobe, T., et al., Mol. Cell Proteomics, 2006, vol. 5, no. 7, p. 1338. 98. Rinner, O., Mueller, L.N., Hubalek, M., Muller, M., et al., Nat. Biotechnol., 2007, vol. 25, no. 3, p. 345. 99. Steen, H., Jebanathirajah, J.A., Springer, M., and Kir schner, M.W., Proc. Natl. Acad. Sci. USA, 2005, vol. 102, no. 11, p. 3948. 100. Rappsilber, J., Ryder, U., Lamond, A.I., and Mann, M., Genome Res., 2002, vol. 12, no. 8, p. 1231. 101. Ishihama, Y., Oda, Y., Tabata, T., Sato, T., et al., Mol. Cell Proteomics, 2005, vol. 4, no. 9, p. 1265. 102. Oda, Y., Huang, K., Cross, F.R., Cowburn, D., and Chait, B.T., Proc. Natl. Acad. Sci. USA, 1999, vol. 96, no. 12, p. 6591. 103. Ong, S.E., Blagoev, B., Kratchmarova, I., Kristensen, D.B., et al., Mol. Cell Proteomics, 2002, vol. 1, no. 5, p. 376. 104. Ibarrola, N., Kalume, D.E., Gronborg, M., Iwahori, A., and Pandey, A., Anal. Chem., 2003, vol. 75, no. 22, p. 6043. 105. Kratchmarova, I., Blagoev, B., HaackSorensen, M., Kas sem, M., and Mann, M., Science, 2005, vol. 308, no. 5727, p. 1472. 106. Blagoev, B., Ong, S.E., Kratchmarova, I., and Mann, M., Nat. Biotechnol., 2004, vol. 22, no. 9, p. 1139. 107. Bose, R., Molina, H., Patterson, A.S., Bitok, J.K., et al., Proc. Natl. Acad. Sci. USA, 2006, vol. 103, no. 26, p. 9773. 108. Goss, V.L., Lee, K.A., Moritz, A., Nardone, J., et al., Blood, 2006, vol. 107, no. 12, p. 4888. 109. Krijgsveld, J., Ketting, R.F., Mahmoudi, T., Johansen, J., et al., Nat. Biotechnol., 2003, vol. 21, no. 8, p. 927. 110. Wu, C.C., MacCoss M.J., Howell K.E., Matthews D.E., Yates J.R. Anal. Chem., 2004, vol. 76, no. 17, p. 4951. 111. Gygi, S.P., Rist, B., Gerber, S.A., Turecek, F., et al., Nat. Biotechnol., 1999, vol. 17, no. 10, p. 994. 112. Yi, E.C., Li, X.J., Cooke, K., Lee, H., et al., Proteomics, 2005, vol. 5, no. 2, p. 380. 113. Ong, S.E., Kratchmarova, I., and Mann, M., J. Proteome Res., 2003, vol. 2, no. 2, p. 173. 114. Thelemann, A., Petti, F., Griffin, G., Iwata, K., et al., Mol. Cell Proteomics, 2005, vol. 4, no. 4, p. 356. No. 13

2010

1310

KAN’SHIN et al.

115. Ross, P.L., Huang, Y.N., Marchese, J.N., Williamson, B., et al., Mol. Cell Proteomics, 2004, vol. 3, no. 12, p. 1154. 116. Munton, R.P., TweedieCullen, R., Livingstone Zatchej, M., Weinandy, F., et al., Mol. Cell Proteomics, 2007, vol. 6, no. 2, p. 283. 117. Zhang, Y., WolfYadlin, A., Ross, P.L., Pappin, D.J., et al., Mol. Cell Proteomics, 2007, vol. 4, no. 9, p. 1240. 118. Schmelzle, K., Kane, S., Gridley, S., Lienhard, G.E., and White, F.M., Diabetes, 2006, vol. 55, no. 8, p. 2171. 119. WolfYadlin, A., Kumar, N., Zhang, Y., Hautaniemi, S., et al., Mol. Syst. Biol, 2006, vol. 2, p. 54. 120. Goshe, M.B., Conrads, T.P., Panisko, E.A., Angell, N.H., et al., Anal. Chem., 2002, vol. 74, no. 3, p. 607.

121. Qian, W.J., Goshe, M.B., Camp, D.G., Yu, L.R., et al., Anal. Chem., 2003, vol. 75, no. 20, p. 5441. 122. Beausoleil, S.A., Villen, J., Gerber, S.A., Rush, J., and Gygi, S.P., Nat. Biotechnol., 2006, vol. 24, no. 10, p. 1285. 123. Hornbeck, P.V., Chabra, I., Kornhauser, J.M., Skrzypek, E., and Zhang, B., Proteomics, 2004, vol. 4, no. 6, p. 1551. 124. http://www.peptideatlas.org/ (Accessed date 26.02.2009). 125. http://www.humanproteinpedia.org/ (Accessed date 26.02.2009). 126. Pedrioli, P.G.A., et al, Nat. Biotechnol., 2004, vol. 22, p. 1459. 127. Martens, L., et al., Proteomics, 2005, vol. 5, p. 3501. 128. http://www.psidev.info/ (Accessed date 26.02.2009).

JOURNAL OF ANALYTICAL CHEMISTRY

Vol. 65

No. 13

2010