proteomics approaches to understand protein phosphorylation in pathway modulation

Available online at www.sciencedirect.com

Proteomics approaches to understand protein phosphorylation inpathway modulationWaltraud X Schulze

Signaling pathways in all organisms consist of series of

phosphorylation and dephosphorylation events that define

directionality and allow different levels of feedback-regulation.

Mass spectrometry-based proteomic analyses in recent years

have led to a proteome-wide identification of thousands of

phosphorylation sites in various plant species. Given this

magnitude of mostly qualitative information about protein

phosphorylation, discovery of specific phosphoproteins with

regulatory functions represents a major challenge. In future

large-scale experiments, combinations of data-driven

modeling strategies based on quantitative data, targeted

kinase–substrate screens, and verification in biochemical and

genetic experiments are required to specifically spot

phosphorylation sites with specific roles in signaling pathway

modulation.

Address

Max Planck Institut fur Molekulare Pflanzenpyhsiologie, Am Muhlenberg

1, 14476 Golm, Germany

Corresponding author: Schulze, Waltraud (wschulze@mpimp-

golm.mpg.de)

Current Opinion in Plant Biology 2010, 13:280–287

This review comes from a themed issue on

Physiology and metabolism

Edited by Uwe Sonnewald and Wolf B. Fromm

Available online 22nd January 2010

1369-5266/$ – see front matter

# 2010 Elsevier Ltd. All rights reserved.

DOI 10.1016/j.pbi.2009.12.008

IntroductionApart from regulation of protein abundance by transcrip-

tion and translation, direct fine-tuned control of protein

activity, protein abundance, or protein localization is

achieved at the post-translational level. Phosphorylation

of serine, threonine, tyrosine, and also of histidine and

aspartate are considered as most important regulatory

post-translational protein modifications in all organisms.

Kinases and phosphatases counterbalance phosphoryl-

ation of their target proteins, thereby achieving specificity

and fine control in signaling pathways through multiple

regulatory feedback loops [1]. Protein kinases make up

about 5.5% of the Arabidopsis genome [2]. This fraction is

nearly twice as high as that in mammals [3], suggesting a

particular complex network and high specificity of phos-

phorylation events in plants. The classic view of signal


transduction pathways leads from receptor proteins at the

plasma membrane to transcription factor proteins in the

nucleus. However, signaling pathways also involve intra-

cellular metabolite receptors transmitting information to

effector enzymes.

Protein phosphorylation often leads to a structural change

of the protein that can directly modulate protein activity,

and induce changes in interaction partners or subcellular

localization. Owing to the tight spatial and temporal

control observed in signaling pathways, and as a means

of achieving directionality in signaling networks, protein

phosphorylation events with regulatory functions are

often of low stoichiometry and transient nature [4]. Phos-

phorylation stoichiometry was found between <10% and

90% depending on the phosphorylation site and protein,

but is rarely being determined on a proteome-wide scale,

as this requires splitting of the sample, dephosphorylation

of one set, and accurate comparative quantitation [5].

Especially in enzymes and kinases phosphorylation sites

with about 5% phosphorylation can trigger functional

effects [5].

In plants, regulation of proteins through phosphorylation

has been studied extensively on purified proteins and by

site-directed mutagenesis of phosphorylation sites.

Especially regulatory phosphorylation on metabolic

enzymes, such as nitrate reductase or sucrose phosphate

synthase were studied thoroughly [6,7]. In recent years,

the identification of protein phosphorylation sites has

become routine through detection of phosphorylated

peptides by mass spectrometry [8]. The breakthrough

in efficient proteome-wide analysis of phosphorylation

sites came with development of suitable enrichment

methods for phosphoproteins or phosphopeptides from

complex protein digests (Table 1). At the same time,

technical advances of mass spectrometer sensitivity to

subfemtomolar detection limits, increased mass accuracy

to less than one ppm, as well as higher resolution and

dynamic range increased confidence in protein identifi-

cation. The development of suitable data-dependent ion

scan procedures [9,10], soft ion fragmentation methods

[11] (Table 2), and identification of typical caveats in

fragmentation spectra interpretation [12] finally resulted

in a boost in the hunt for protein phosphorylation sites.

This review covers large-scale datasets collected over the

past three years and addresses the need for future exper-

imental and computational strategies to better under-

stand the biological role of protein phosphorylation in

signaling pathways on a proteome-wide scale.

www.sciencedirect.com

mailto:[email protected]

mailto:[email protected]

http://dx.doi.org/10.1016/j.pbi.2009.12.008

Proteomics approaches to understand protein phosphorylation in pathway modulation Schulze 281

Table 1

Methods of phosphopeptide enrichment are biased toward selective phosphoproteome subsets. Overlap of identified phosphopeptides

between two different methods can be as low as 30% [57��]. For best phosphopeptide coverage, more than one method need to be used

Resin type Metal ion Efficiency

(%)

Quenching agent Comments Examples

Affinity purification

IMAC Fe3+, Ga3+,

Zn2+50–90 Sometimes chemical

derivatization

(methyl-esterification)

Non-specific binding of acidic peptides to

the matrix. Specificity can be increased

through derivatization, but this is often

involved with general loss of sample.

More efficiently for multiply-phosphorylated peptides.

[14,15�,

23�,24�]

TiO2 Ti2+ 30–60 2,5-Dihydroxybenzoic

acid

Non-specific binding of acidic peptides can be

quenched by different acidic chemicals.

More efficiently for singly phosphorylated peptides.

[57��]

60–80 Phthalic acid [15�]

60–80 Lactic acid [14]

ZrO2 Zr2+ 60 ß-Hydroxypropanoic

acid

[57��]

Al(OH)3 Al3+ 30 Can be used also to purify phosphoproteins. [58]

Chemical purification

Phosphoramidate

chemistry (PAC)

70 Phosphopeptides are coupled to solid-phase matrix;

elution under acidic conditions. Phosphate group

remains attached to peptide, efficient for singly

phosphorylated peptides.

[57��]

Beta elimination 80–90 Resulting double bond can react with nucleophilic

functional groups for selective purification.

Possible side reaction. Phosphate group is

removed from peptide.

[59]

Large-scale mass spectrometry-basedphosphoproteomics datasetsQualitative datasets

Numerous plant phosphoproteomics studies have been

published identifying hundreds to thousands of phos-

phorylation sites in various plant species [13–27]. In all

of these phosphoproteomic profiling studies, phospho-

peptide enrichment using metal oxide chromatography

has been applied, and in most studies prefractionation to

specific organelles (chloroplast, tonoplast, nucleus, and

plasma membrane), or peptide fractionation by ion

exchange chromatography was carried out. Among the

studies focusing on Arabidopsis, 6% of all identified

phosphorylation sites were found by more than two

independent experiments, but 83% of the published

phosphorylation sites were identified only once [28].

Despite impressive advances in identification of thou-

sands of plant phosphorylation sites, they probably still

represent a rather incomplete subset of the entire phos-

phoproteome. It is not yet clear, how many of the about

two million potentially phosphorylatable sites in Arabi-

dopsis (1 142 488 S, 664 750 T, 369 122 Y) are actually invivo being accessible to modification and are then used

under specific conditions. A high confident positive phos-

phorylation site prediction was obtained for nearly

500 000 of these residues (203 622 S, 174 301 T, and

120 983 Y), and experimental evidence from biochemistry

or mass spectrometry was yet obtained for about 12 000


residues (9406 S, 2352 T, 699 Y) covering about 5000

proteins in Arabidopsis [28]. Also, it remains open to

which extent phosphoproteome compositions will vary

among the many different cell and tissue types in multi-

cellular organisms [29]. Although different growth con-

ditions and tissues have been analyzed [28], functional

conclusions based on qualitative data are yet difficult to

draw. This is due to biases in isolated phosphoproteomes

by different enrichment methods and due to differences

in phosphopeptide identification depending on instru-

mentation used (Tables 1 and 2). In addition, key regu-

latory proteins such as signaling proteins and transcription

factors are often of low cellular abundance [30].

The increasing qualitative information about protein

phosphorylation in various databases (Table 3), has led

to a disconnection from functional aspects, as the majority

of sites remains uncharacterized. It has in fact been

suggested that large numbers of the phosphorylation sites

identified by proteomic methods could be non-functional

[31] making an important question apparent: What are the

appropriate screens in finding phosphorylation sites

which have regulatory roles? Given the difficulties

described above, detecting ‘relevant’ phosphorylation

sites by unbiased qualitative profiling methods may not

be so straight forward. Nevertheless, simple experimental

evidence of specific phosphorylation sites has in many

cases helped defining experimental targets. For example,

T881 was confirmed as additional regulatory site in the


282 Physiology and metabolism

Table 2

Mass spectrometric fragmentation methods to study phosphopeptides. Phosphoserine (pS) and phosphothreonine (pT) sites are prone to

neutral loss of phosphoric acid during high-energy peptide fragmentation and are then recognized by the presence of fragment ions

containing dehydroalanine (pS) and dehydrobutyric acid (pT). Phosphotyrosine (pY) residues rarely undergo neutral loss. The advantages

of neutral-loss driven MS3 and MSA scans in large-scale phosphoproteomics are controversial and may depend on instrumentation [9,60]

Fragmentation

type

Instrument Properties Comments Examples

CID MS2 Ion trap

Q-TOF

LTQ-Orbitrap

Often considered ineffective for

phosphopeptide analysis as

neutral loss of phosphoric acid can

occur before peptide fragmentation.

Thus, insufficient fragment ions for

identification of peptide sequence.

MSA and MS3 scans are generally

capable of producing information rich

spectra, but for high mass accuracy instruments

with large ion capacity, CID MS2 scans are

better suited for phosphopeptide

identification than initially anticipated.

MS3-based scans can complicate analysis

through neutral losses of methanesulfonic

acid from methionine.

[14,15�,23�]

CID MS3 Ion trap

LTQ-Orbitrap

Peak selected for additional

fragmentation is

isolated before activation, resulting in new

set of product ions. MS3 scans are

20% slower and hold only 15%

of the ion intensity of MS2 spectra.

[19,24�]

CID MSA Ion trap

LTQ-Orbitrap

Second isolation step is eliminated and

spectrum contains product ions from

the original peptide ion as well as from

the neutral-loss activation event.

A pseudo-MS3 approach.

[61]

HQD LTQ-Orbitrap Information rich CID spectra without

low-mass cut-off.

Suitable to use in

combination with

iTRAQ quantification.

[10]

ETD Ion trap

LTQ-Orbitrap

Minimal neutral loss, fast enough for

chromatographic time scale.

Implementation of soft fragmentation

methods without neutral-loss aids

localization of phosphorylation

site within the peptide sequence.

[16�]

ECD

IRMPD

FT-ICR Minimal neutral loss, too slow for

chromatographic time scale.

n.a.

CID: collision induced dissociation; HQD: higher collision decomposition device; ETD: electron transfer dissociation; ECD: electron capture

dissociation; IRMPD: infrared multiphoton dissociation. MS2: fragment spectrum after dissociation of a precursor peptide; MS3: fragment spectrum

after further dissociation of a fragment ion from the MS2 scan, usually the neutral-loss peak; and MSA: multistage activation.

Table 3

Resources about protein phosphorylation and pathway context

Database Weblink Organisms Reference Features

Plant phosphorylation site databases

P3DB http://digbio.missouri.edu/p3db/ All plants, focus on Brassica

napus

[62] BLAST tool

PhosPhAt http://phosphat.mpimp-golm.mpg.de Plants, focus on Arabidopsis; [28] Predictor pS, pT, pY; motif

search; functional classification;

experimental context information

Medicago http://phospho.medicago.wisc.edu/ Medicago trunculata [16�] BLAST tool; motif search

Non-plant phosphorylation site databases

Phosida http://www.phosida.org Human, mouse, C. elegans,

Drosophila, yeast, prokaryota,

archaea;

[63] Temporal information; motif

search; protein context information

PhosphoELM http://phospho.elm.eu.org Eukaryota [64] Probably most complete

dataset for many eukaryotes

PhosphoPEP http://www.phosphopep.org Yeast, C. elegans, Drosophila,

human

[65] Pathway links

Resources with kinase–substrate information and network reconstruction

NetworKIN http://networkin.info/search.php Human, C. elegans,

Drosophila, yeast

[66] Predicted pathways

PhosphoPOINT http://kinase.bioinformatics.tw/ Human [54] Predicted pathways

Current Opinion in Plant Biology 2010, 13:280–287 www.sciencedirect.com

http://digbio.missouri.edu/p3db/

http://phosphat.mpimp-golm.mpg.de/

http://phospho.medicago.wisc.edu/

http://www.phosida.org/

http://phospho.elm.eu.org/

http://www.phosphopep.org/

http://networkin.info/search.php

http://kinase.bioinformatics.tw/


plasma membrane ATPase [24�], or the regulatory role of

T464 in activation of ammonium transporter AMT1;1

[32] was analyzed after this site was published in a

systematic study [13].

Quantitative datasets

One strategy in systematic identification of phosphoryl-

ation sites with regulatory significance is the comparison

of phosphorylation levels under differential conditions.

This allows distinguishing condition-specific responses

among a majority of condition-independent phosphoryl-

ation events. Similarly as genome-wide co-expression

analyses have become a key tool in identification of genes

involved in a biological pathway and in annotation of

unknowns [33], comparative analysis of quantitative site-

specific phosphorylation profiles under various conditions

are likely to contribute to our understanding of their

relevance in pathway control.

The extraction of reliable quantitative data from phos-

phoproteomics experiments requires careful experimen-

tal design regarding suitable choice of quantitation

method, randomized blocking of treatments and repli-

cates, and definition of controls. In plants, comparative

studies of individual phosphopeptides have been carried

out mainly under two different conditions, such as elicitor

treatment versus control [21�], mutant versus wild type

[22] or analyzing the effect of various stresses [20,25,27].

A study in chloroplasts compared the phosphorylation

levels of casein kinase targets from day and night [15�],and a quantitative study of protein phosphorylation

changes in response to brassinosteroid treatment revealed

new BRI1 signaling candidate proteins [34].

The principal challenge in carrying out comprehensive

peptide-based quantitative phosphoproteomic analyses

lies in efficient phosphopeptide detection with deep

proteome coverage in combination with accurate quanti-

fication strategies. The most comprehensive study of site-

specific phosphorylation dynamics to date comes from a

five point time course of EGF growth factor stimulation in

mammalian HeLa cells [35]. The acquired phosphoryl-

ation profiles were assigned to clusters that were differ-

entially enriched for related signaling components

involved at different phases of signal transduction. In

plants, two similar experiments have been done using

iTRAQ labeling and label-free quantitation to study

time-resolved phosphorylation changes after treatment

of cell cultures with the elicitor peptide flg22 [23�] or to

study the response of seedlings to supply of external

sucrose [24�]. However, in both cases, the number of

quantified phosphorylation sites was rather low.

In quantitative datasets, phosphopeptide abundance

change needs to be normalized for changes in whole

protein abundance by using quantitative information of

non-phosphopeptides of the same protein to determine


whether phosphopeptide changes are due to real phos-

phorylation changes, or whether they instead result from

altered protein abundance [36]. As protein degradation

and spatial translocation can be induced rapidly, and

protein abundance changes resulting from transcriptional

regulation can occur within 30 min after treatment, the

parallel quantitative acquisition of phosphopeptide and

non-phosphopeptide datasets is highly relevant for func-

tional analyses.

In all phosphopeptide-based analyses, many of the ident-

ified phosphoproteins contained more than one phos-

phorylation site that was often found to be

differentially affected by the treatment [23�,24�,35]. This

emphasizes the importance of site-specific quantitative

analysis. Among the plant proteins with more than two

phosphorylation sites, protein kinases (GO:0004672) and

proteins with nucleotide binding activity (GO:0000166)

are over-represented suggesting that protein kinases and

transcription factors are important proteins of regulation

and integration in cellular signaling pathways [28]. How-

ever, multiple phosphorylation sites per protein compli-

cate the identification of sites with strong functional roles,

as the ‘relevance’ of specific phosphorylation sites may be

specific to tissue types or condition-dependent [29].

Thus, future phosphoproteomic experiments will even

more have to be of quantitative nature in order to progress

from phosphorylation site identification to a functional

characterization (Figure 1).

Targeted analysis of kinase and substraterelationshipsIn order to infer the topology of signaling networks from

large-scale phosphoproteomic data, knowledge about the

kinase–substrate interactions is necessary. Yeast two-

hybrid screens, genetic techniques as well as biochemical

purification strategies identified individual kinase–sub-

strate pairs or verified their in vivo relevance. However,

most of these approaches have limited throughput when

applied on proteome-wide scale, and they are often not

readily applicable to modification-dependent responses.

Thus, there still is a large gap between identification of invivo phosphorylation sites and the kinases that modulate

them. Despite computational approaches such as consen-

sus motif identification [37], for about 18 000 annotate

phosphorylation sites in the Phospho.ELM database, only

25% have been linked to at least one in vivo kinase [4].

An experimental strategy for systematic large-scale

identification of kinase targets involves screening for

phosphorylation of short linear sequence motifs either

using custom selection of experimentally verified target

peptides or degenerate peptide sets (Figure 1). Semi-

degenerate peptide arrays have been used to screen for

phosphorylation preferences of members of the CDPK

and SnRK kinase family [38��]. Also, peptide arrays

designed from in vivo identified phosphorylation sites



Figure 1

Workflow from identification of phosphorylation sites to characterization of the functional role in a pathway context. Large-scale untargeted

experiments need to be refined in focused functional studies and supported by computational models. Specific candidate phosphoproteins or

phosphorylation sites finally need in planta validation.

can give insights into regulatory roles of chosen phospho-

peptides [25]. Similarly, kinase-specific phosphorylation

of proteins immobilized to arrays can be analyzed in high-

throughput fashion: By testing 87 out of 122 yeast protein

kinases individually for phosphorylation of 4400 immobil-

ized proteins, each kinase had 47 protein targets on

average, and about one third of the spotted proteins were

targeted at least once [39]. Related protein kinases phos-

phorylated different sets of proteins, suggesting that

protein and peptide arrays can be useful tools in the

identification of specific protein kinase substrates. In a

comprehensive study in plants using the MAPK pathway

as an example, the activation preferences of different

MAPKKs were determined and a subset of activated

MAPKs were applied to a high-density protein microarray

to determine their phosphorylation targets [40��]. How-

ever, the in vitro array experiments study kinase–sub-

strate relationships disconnected from the in vivosignaling context.

A complementary strategy for targeted phosphorylation

analysis is by focused analysis of specific ‘reporter’ phos-

phorylation sites from proteins of interest using selected

reaction monitoring (SRM). In this method, a triple-


quadrupole mass spectrometer is used to specifically

isolate the target peptide ion of interest in the first

quadrupole. Upon fragmentation of this ion in the second

quadrupole, characteristic fragment ions are specifically

detected in the third quadrupole with very low noise

levels. This approach has been pioneered to study phos-

phorylation of isoforms of trehalose phosphate synthase

[41] and is likely to gain importance in future to carry out

targeted quantitative comparative analyses of specific

phosphopeptides.

Phosphorylation in the signaling pathwaycontextIn most plant signaling networks, our view of complete

pathways from stimulus to response is still limited. The

MAP-kinase pathway is among the most well studied

central signaling pathway with functions in biotic stress

responses, plant defense, hormone signaling, senescence

and development, as well as a number of abiotic stresses

(reviewed in Refs. [42,43]). With regards to metabolic

signaling, in the recent years the Snf1-related protein

kinases (SnRKs) became apparent as central players link-

ing stress responses and metabolic signaling [44–46].

Calcium-dependent protein kinases have been studied



intensively as a plant specific protein kinase family

involved in biotic and abiotic stress responses [47]. Less

well understood is yet the role of the CBL–CIPK kinase

network, which is involved in regulation of ion transport

across membranes (reviewed in Ref. [48]). Although phos-

phorylation sites in the activation loops have been ident-

ified for many kinases of the abovementioned signaling

pathways [28], their dynamics of activation and inactivation

by phosphorylation is only starting to be under systematic

investigation.

The majority of proteins in the largest plant kinase

family, the receptor like kinases [49] are still uncharac-

terized. However, the flagellin receptor FLS2 and the

brassinosteroid receptor BRI1 connect to well character-

ized signaling cascades reaching from the membrane to

the nucleus [50��,51]. Untargeted qualitative and quan-

titative identification of phosphorylation sites has sub-

stantially contributed to the identification of new target

candidates in these pathways [23�,34,50��]. In case of the

BRI1 signaling pathway, combination of quantitative

brassinolide-dependent identification of phosphorylation

changes on 2D-gels with targeted site-directed mutagen-

esis of phosphorylation sites and analysis of suitable

knockout mutants has led to complete elucidation of

the signaling pathway from autophosphorylation of re-

ceptor BRI1 to the specific activation of transcription by

BZR1. Phosphorylation in this pathway not only regulates

activity of kinases BRI1, BSK2 and BIN1, but also modu-

lates subcellular localization of BZR1 [50��].

Adding another layer of complexity, signaling networks

involve not only protein modification through phos-

phorylation, but also redox-regulation and ubiquitinyla-

tion. Often these converge on the same target proteins

and influence each other. Phosphorylation was shown to

be inhibited by methionine oxidation in near phosphoryl-

ation sites for CDPK recognition motifs and on a regu-

latory phosphorylation site of nitrate reductase [52].

Data-derived identification of regulatoryphosphorylation sitesLinear motifs around transient phosphorylation sites in

combination with interaction specificity of protein domains

are structural features of the modular nature of protein

signaling pathways. Characterization of such modification-

dependent regulatory interaction networks on a proteome-

wide scale requires extensive computational analysis and

data-derived modeling in addition to well-designed exper-

iments acquiring quantitative data [4].

Large-scale mass spectrometry-based phosphoproteomic

approaches in combination with systematic analysis of

kinase targets as well as the exploitation of information

from general protein–protein interaction studies have

resulted in prediction models for kinase–substrate

relationships and phosphorylation networks for a variety


of organisms [53��,54]. These approaches are promising as

hypothesis-generating platforms and are indeed also

applicable to plant models, for which co-expression

analysis platforms already exist (reviewed in Ref. [33]),

information of subcellular localization is available [55],

and a predicted interactome has been published [56]. The

integration of functional genomic and phosphoproteomic

data represents a powerful strategy for constructing com-

prehensive maps of signaling networks and predicting the

functional roles of their individual components on a

proteome-wide level (Figure 1).

ConclusionExperimental strategies to analyze protein phosphoryl-

ation in signaling pathways in future will have to combine

untargeted large-scale qualitative and quantitative mass

spectrometry-based phosphopeptide identifications with

equally important refinement in targeted experiments.

These can imply selected reaction monitoring exper-

iments using proteotypic phosphopeptide standards, or

systematic kinase–substrate screens involving peptide or

protein microarrays. Regarding the functional analysis of

individual phosphorylation sites, data-driven compu-

tational modeling will probably gain importance to gen-

erate hypotheses and identify putative experimental

targets. However, finally it still leaves researchers with

the difficult task of experimentally validating candidate

kinase–substrate pairs or with testing the biological

relevance of protein phosphorylation by low-throughput

analysis of mutants with altered expression levels and

mutated phosphorylation sites.

AcknowledgementsI am very thankful to Birgit Kersten, Annemarie Matthes, WolfgangEngelsberger, Sylwia Kierszniowska and Alois Schweighofer for criticalcomments on the manuscript and fruitful discussions. Birgit Kersten,Wolfgang Engelsberger and Robert Schmidt provided experimental orcomputational background information.

References and recommended readingPapers of particular interest, published within the period of review,have been highlighted as:

� of special interest�� of outstanding interest

1. Brandman O, Meyer T: Feedback loops shape cellular signals inspace and time. Science 2008, 322:390-395.

2. Initiative TAG: Analysis of the genome sequence of theflowering plant Arabidopsis thaliana. Nature 2000, 408:796-816.

3. Manning G, Whyte DB, Martinez R, Hunter R, Hunter T,Sudarsanam S: The protein kinase complement of the humangenome. Science 2002, 298:1912-1914.

4. Jorgensen C, Linding R: Directional and quantitativephosphorylation networks. Brief Functional Genomics andProteomics 2008, 7:17-26.

5. Mayya V, Han DK: Phosphoproteomics by mass spectrometry:insights, implications, applications and limitations. ExpertReview of Proteomics 2009, 6:605-618.

6. Toroser D, Athwal GS, Huber SC: Site-specific regulatoryinteraction between spinach leaf sucrose-phosphatesynthase and 14-3-3 proteins. FEBS Letters 1998, 435:110-114.



7. Kaiser WM, Huber SC: Post-translational regulation of nitratereductase: mechanism, physiological relevance andenvironmental triggers. Journal of Experimental Botany 2001,52:1981-1989.

8. Steen H, Jebanathirajah JA, Rush J, Morrice N, Kirschner MW:Phosphorylation analysis by mass spectrometry: myths, factsand the consequence for qualitative and quantitativemeasurements. Molecular and Cellular Proteomics 2006,5:172-181.

9. Villen J, Beausoleil SA, Gygi SP: Evaluation of the utility ofneutral-loss-dependent MS3 strategies in large-scalephosphorylation analysis. Proteomics 2008, 8:4444-4452.

10. Zhang Y, Ficarro SB, Li S, Marto JA: Optimized Orbitrap HCD forquantitative analysis of phosphopeptides. Journal of theAmerican Society of Mass Spectrometry 2009, 20:1425-1434.

11. Molina H, Horn DM, Tang N, Mathivanan S, Pandey A: Globalproteomic profiling of phosphopeptides using electrontransfer dissociation tandem mass spectrometry. InProceedings of the National Academy of Sciences of the UnitedStates of America 2007, 104:2199-2204.

12. Lehmann WD, Kruger R, Salek M, Hung CW, Wolschin F,Weckwerth W: Neutral loss-based phosphopeptiderecognition: a collection of caveats. Journal of ProteomeResearch 2007, 6:2866-2873.

13. Nuhse TS, Stensballe A, Jensen ON, Peck SC:Phosphoproteomics of the Arabidopsis plasma membraneand a new phosphorylation site database. Plant Cell 2004,16:2394-2405.

14. Sugiyama N, Nakagami H, Mochida K, Daudi A, Tomita M,Shirasu K, Ishihama Y: Large-scale phosphorylation mappingreveals the extent of tyrosine phosphorylation in Arabidopsis.Molecular Systems Biology 2008, 4:e1-7.

15.�

Reiland S, Messerli G, Baerenfaller K, Gerrits B, Endler A,Grossmann J, Gruissem W, Baginsky S: Large-scale Arabidopsisphosphoproteome profiling reveals novel chloroplast kinasesubstrates and phosphorylation networks. Plant Physiology2009, 150:889-903.

Most comprehensive phosphoproteomic analyses of plant chloroplast.Reconstruction of kinase–substrate relationships based on motif analysisof sequences around phosphorylation sites using the casein kinase II asan example. Comparison of phosphorylation status of key target proteinsbetween light and dark conditions.

16.�

Grimsrud PA, den Os D, Wenger CD, Swaney DL, Schwartz D,Sussman MR, Ane JM, Coon JJ: Large-scale phosphoproteinanalysis in Medicago truncatula roots provides insight into invivo kinase activity in legumes. Plant Physiology 2009, 152:19-28.

Large-scale identification of phosphorylation sites in a legume using ETDfragmentation in the mass spectrometric analysis. Motif analysis andcross-species comparison of phosphorylation sites.

17. Tan F, Li G, Chitteti BR, Peng Z: Proteome andphosphoproteome analysis of chromatin associated proteinsin rice (Oryza sativa). Proteomics 2007, 7:4511-4527.

18. Jones AM, Maclean D, Studholme DJ, Serna-Sanz A,Andreasson E, Rathjen JP, Peck SC: Phosphoproteomicanalysis of nuclei-enriched fractions from Arabidopsisthaliana. Journal of Proteomics 2009, 72:439-451.

19. Whiteman S-A, Serazetdinova L, Jones AM, Sanders D, Rathjen J,Peck SC, Maathuis FJM: Identification of novel proteins andphosphorylation sites in a tonoplast enriched membranefraction of Arabidopsis thaliana. Proteomics 2008, 8:3536-3547.

20. Schneider T, Schellenberg M, Meyer S, Keller F, Gehrig P,Riedel K, Lee Y, Eberl L, Martinoia E: Quantitative detection ofchanges in the leaf-mesophyll tonoplast proteome independency of a cadmium exposure of barley (Hordeumvulgare L.) plants. Proteomics 2009, 9:2668-2677.

21.�

Benschop JJ, Mohammed S, O’Flaherty M, Heck AJ, Slijper M,Menke FL: Quantitative phospho-proteomics of early elicitorsignalling in Arabidopsis. Molecular and Cellular Proteomics2007, 6:1705-1713.

Elegant quantitative phosphoproteomic profiling in Arabidopsis cell cul-tures comparing elicitor treated cells with controls using stable isotopelabeling and reciprocal experimental design. This allows precise compar-


ison of phosphorylation levels between treatments. The work providesinsights into specificity and divergence of two elicitor responses.

22. Li H, Wong WS, Zhu L, Guo HW, Ecker J, Li N: Phosphoproteomicanalysis of ethylene-regulated protein phosphorylation inetiolated seedlings of Arabidopsis mutant ein2 using two-dimensionalseparationscoupledwithahybridquadrupole time-of-flight mass spectrometer. Proteomics 2009, 9:1646-1661.

23.�

Nuhse TS, Bottrill AR, Jones AM, Peck SC: Quantitativephosphoproteomic analysis of plasma membrane proteinsreveals regulatory mechanisms of plant innate immuneresponses. Plant Journal 2007, 51:931-940.

Most comprehensive phosphoproteomic analyses of plant chloroplast.Reconstruction of kinase–substrate relationships based on motif analysisof sequences around phosphorylation sites using the casein kinase II asan example. Comparison of phosphorylation status of key target proteinsbetween light and dark conditions.

24.�

Niittyla T, Fuglsang AT, Palmgren MG, Frommer WB, Schulze WX:Temporal analysis of sucrose-induced phosphorylationchanges in plasma membrane proteins of Arabidopsis.Molecular and Cellular Proteomics 2007, 6:1711-1726.

One of the first studies of phosphorylation site dynamics over more thantwo conditions in Arabidopsis. Label-free quantitation was used to studythe response of carbon-starved seedlings to sucrose. The known phos-phorylation sites of the plasma membrane ATPase was used as anexample to verify the approach taken and new phosphorylation site ofplasma membrane ATPase was characterized.

25. de la Fuente van Bentem S, Anrather D, Dohnal I, Roitinger E,Csaszar E, Joore J, Buijnink J, Carreri A, Forzani C, Lorkovic ZJet al.: Site-specific phosphorylation profiling of Arabidopsisproteins by mass spectrometry and peptide chip analysis.Journal of Proteome Research 2008, 7:2458-2470.

26. de la Fuente van Bentem S, Anrather D, Roitinger E, Djamei A,Hufnagl T, Barta A, Csaszar E, Dohnal I, Lecourieux D, Hirt H:Phosphoproteomics reveals extensive in vivo phosphorylationof Arabidopsis proteins involved in RNA metabolism. NucleicAcids Research 2006, 34:3267-3278.

27. Stulemejer IJ, Joosten MH, Jensen ON: Quantitativephosphoproteomics of tomato mounting a hypersensitiveresponse reveals a swift suppression of photosyntheticactivity and a differential role for hsp90 isoforms. Journal ofProteome Research 2009, 8:1168-1182.

28. Durek P, Schmidt R, Heazlewood JL, Jones A, MacLean D,Nagel A, Kersten B, Schulze WX: PhosPhAt: The Arabidopsisthaliana phosphorylation site database. An update. NucleicAcids Research 2010, 38:D828-D834.

29. Ubersax JA, Ferrell JEJ: Mechanisms of specificity in proteinphosphorylation. Nature Reviews Molecular Cell Biology 2007,8:530-541.

30. Piques MC, Schulze WX, Hohne M, Gibon Y, Rohwer J, Stitt M:Ribosome and transcript copy numbers, polysome occupancyand enzyme dynamics in Arabidopsis. Molecular SystemsBiology 2009, 5:E1-E17.

31. Lienhard GE: Non-functional phosphorylations? Trends inBiochemical Science 2008, 33:351-352.

32. Lanquar V, Loque D, Hormann F, Yuan L, Bohner A,Engelsberger WR, Lalonde S, Schulze WX, Von Wiren N,Frommer WB: Feedback inhibition of ammonium uptake by aphospho-dependent allosteric mechanism in Arabidopsis. ThePlant Cell 2009, 21:3610-3622.

33. Usadel B, Obayashi T, Mutwil M, Giorgi FM, Bassel GW,Tanimoto M, Chow A, Steinhauser D, Persson S, Provart NJ: Co-expression tools for plant biology: opportunities forhypothesis generation and caveats. Plant, Cell and Environment2009, 32:1633-1651.

34. Tang W, Deng Z, Oses-Prieto JA, Suzuki N, Zhu S, Zhang X,Burlingame AL, Wang ZY: Proteomics studies ofbrassinosteroid signal transduction using prefractionationand two-dimensional DIGE. Molecular and Cellular Proteomics2008, 7:728-738.

35. Olsen JV, Blagoev B, Gnad F, Macek B, Kumar C, Mortensen P,Mann M: Global, in vivo, and site-specific phosphorylationdynamics in signaling networks. Cell 2006, 127:635-648.



36. Steen H, Jebanathirajah JA, Springer M, Kirschner MW: Stableisotope-free relative and absolute quantitation of proteinphosphorylation stoichiometry by MS. In Proceedings of theNational Academy of Sciences of the United States of America2005, 102:3948-3953.

37. Schwartz D, Gygi SP: An iterative statistical approach to theidentification of protein phosphorylation motifs from large-scale data sets. Nature Biotechnology 2005, 23:1391-1398.

38.��

Vlad F, Turk BE, Peynot P, Leung J, Merlot S: A versatile strategyto define the phosphorylation preferences of plant proteinkinases and screen for putative substrates. Plant Journal 2008,55:104-117.

Impressive conceptual study of the use of 198 semi-degenerate peptidespottend on arrays to screen for phosphorylation target preferences ofisolated SnRK and CDPK kinases. Phosphoimaging was used for quan-titation of the implementation of radioactive ATP. Quantitative data werethen converted into position-specific scoring matrices to identify putativesubstrates of these kinases in silico from protein sequence databases.

39. Ptacek J, Devgan G, Michaud G, Zhu H, Zhu X, Fasolo J, Guo H,Jona G, Breitkreutz A, Sopko R et al.: Global analysis of proteinphosphorylation in yeast. Nature 2005, 438:679-684.

40.��

Popescu SC, Popescu GV, Bachan S, Zhang Z, Gerstein M,Snyder M, Dinesh-Kumar SP: MAPK target networks inArabidopsis thaliana revealed using functional proteinmicroarrays. Genes and Development 2009, 23:80-92.

Systematic and comprehensive analysis of MKK and MPK targets usingprotein microarrays. In a first step, MKK activation targets were identified,and activated MPKs were then tested for their substrates on a high-density protein microarray. Selected kinase–substrate relationships werevalidated in planta.

41. Glinski M, Weckwerth W: Differential multisite phosphorylationof the trehalose-6-phosphate synthase gene family inArabidopsis thaliana: A mass spectrometry-based process formultiparallel peptide library phosphorylation analysis.Molecular and Cellular Proteomics 2005, 4:1614-1625.

42. Fiil BK, Petersen K, Petersen M, Mundy J: Gene regulation byMAP kinase cascades. Current Opinion in Plant Biology 2009,12:615-621.

43. Pitzschke A, Schikora A, Hirt H: MAPK cascade signallingnetworks in plant defence. Current Opinion in Plant Biology 2009,12:421-426.

44. Halford NG, Hey SJ: Snf1-related protein kinases (SnRKs) actwithin an intricate network that links metabolic and stresssignalling in plants. Biochemical Journal 2009, 419:247-259.

45. Shin R, Alvarez S, Burch AY, Jez JM, Schachtman DP:Phosphoproteomic identification of targets of the Arabidopsissucrose nonfermenting-like kinase SnRK2.8 reveals aconnection to metabolic processes. In Proceedings of theNational Academy of Sciences of the United States of America2007, 104:6460-6465.

46. Fragoso S, Espındola L, Paez-Valencia J, Gamboa A, Camacho Y,Martınez-Barajas E, Coello P: SnRK1 isoforms AKIN10 andAKIN11 are differentially regulated in Arabidopsis plants underphosphate starvation. Plant Physiology 2009, 149:1906-1916.

47. Ludwig AA, Romeis T, Jones JD: CDPK-mediated signallingpathways: specificity and cross-talk. Journal of ExperimentalBotany 2004, 55:181-188.

48. Weinl S, Kudla J: The CBL-CIPK Ca(2+)-decoding signalingnetwork: function and perspectives. New Phytologist 2009,2009:3.

49. Shiu S-H, Bleecker AB: Plant receptor-like kinase gene family:diversity, function, andsignaling. Science’s STKE 2001,113:1-13.

50.��

Kim TW, Guan S, Sun Y, Deng Z, Tang W, Shang JX, Sun Y,Burlingam A, Wang ZY: Brassinosteroid signal transductionfrom cell-surface receptor kinases to nuclear transcriptionfactors. Nature Cell Biology 2009, 11:1254-1260.

Excellent example of how screening approaches for quantitative phos-phorylation responses can contribute to identify missing components in asignaling pathway. Candidate proteins were subjected to tedious indivi-dual characterization involving site-directed mutagenesis of phosphor-ylation sites and genetic analysis of plant mutants.


51. Chinchilla D, Boller T, Robatzek S: Flagellin signalling in plantimmunity. Advances in Experimental Medicine and Biology 2007,598:358-371.

52. Hardin SC, Larue CT, Oh MH, Jain V, Huber SC: Couplingoxidative signals to protein phosphorylation via methionineoxidation in Arabidopsis. Biochemical Journal 2009,422:305-312.

53.��

Linding R, Jensen LJ, Ostheimer GJ, van Vugt MA, Jorgensen C,Miron IM, Diella F, Colwill K, Taylor L, Elder K et al.: Systematicdiscovery of in vivo phosphorylation networks. Cell 2007,129:1415-1426.

Combining experimental data and computational methods in the algo-rithm NetworKIN that predicts individual kinases responsible for in vivophosphorylation sites. The NetworKIN algorithm combines evidence fromconsensus recognition motifs, co-expression, localization, or proteinscaffolds.

54. Yang CY, Chang CH, Yu YL, Lin TC, Lee SA, Yen CC, Yang JM,Lai JM, Hong YR, Tseng TL et al.: PhosphoPOINT: acomprehensive human kinase interactome and phospho-protein database. Bioinformatics 2008, 24:14-20.

55. Heazlewood JL, Verboom RE, Tonti-Filippini J, Small I, Millar AH:SUBA: the Arabidopsis subcellular database. Nucleic AcidsResearch 2007, 35:D213-D218.

56. Geisler-Lee J, O’Toole N, Ammar R, Provart NJ, Millar AH,Geisler M: A predicted interactome for Arabidopsis. PlantPhysiology 2007, 145:317-329.

57.��

Bodenmiller B, Mueller LN, Mueller M, Domon B, Aebersold R:Reproducible isolation of distinct, overlapping segments ofthe phosphoproteome. Nature Methods 2007, 4:231-237.

Systematic study of the influence of phosphopeptide enrichment methodon the types of peptides identified by mass spectrometry. The studyprovides quantitative evidence for the degree of methodological bias inphosphoproteomic studies.

58. Wolschin F, Lehmann U, Glinski M, Weckwerth W: An integratedstrategy for identification and relative quantification of site-specific protein phosphorylation using liquid chromatographycoupled to MS2/MS3. Rapid Communications in MassSpectrometry 2005, 19:3626-3632.

59. Hegeman AD, Harms AC, Sussman MR, Bunner AE, Harper JF: Anisotope labeling strategy for quantifying the degree ofphosphorylation at multiple sites in proteins. Journal of theAmerican Society of Mass Spectrometry 2004, 15:647-653.

60. Ulintz PJ, Yocum AK, Bodenmiller B, Aebersold R, Andrews PC,Nesvizhskii A: Comparison of MS(2)-only, MSA, and MS(2)/MS(3) methodologies for phosphopeptide identification.Journal of Proteome Research 2009, 8:887-899.

61. Schroeder MJ, Shabanowitz J, Schwartz JC, Hunt DF, Coon JJ: Aneutral loss activation method for improved phosphopeptidesequence analysis by quadrupole ion trap mass spectrometry.Analytical Chemistry 2004, 76:3590-3598.

62. Gao J, Agrawal GK, Thelen JJ, Xu D: P3DB: a plant proteinphosphorylation database. Nucleic Acids Research 2009,37:D960-D962.

63. Gnad F, Ren S, Cox J, Olsen JV, Macek B, Oroshi M, Mann M:PHOSIDA (phosphorylation site database): management,structural and evolutionary investigation, and prediction ofphosphosites. Genome Biology 2007, 8:R250.

64. Gould CM, Diella F, Via A, Puntervoll P, Gemund C, Chabanis-Davidson S, Michael S, Sayadi A, Bryne JC, Chica C et al.: ELM:the status of the 2010 eukaryotic linear motif resource. NucleicAcids Research 2010, 38:D167-D180.

65. Bodenmiller B, Malmstrom J, Gerrits B, Campbell D, Lam H,Schmidt A, Rinner O, Muller LN, Shannon PT, Pedrioli PG et al.:PhosphoPep — a phosphoproteome resource for systemsbiology research in Drosophila Kc167 cells. Molecular SystemsBiology 2007, 3:139-149.

66. Linding R, Jensen LJ, Pasculescu A, Olhovsky M, Colwill K, Bork P,Yaffe MB, Pawson T: NetworKIN: a resource for exploringcellular phosphorylation networks. Nucleic Acids Research2007, 36:D695-696.


proteomics approaches to understand protein phosphorylation in pathway modulation

Documents