High-sensitivity analysis of specific peptides in complex samples by selected MS/MS ion monitoring...

JOURNAL OF MASS SPECTROMETRYJ. Mass Spectrom. 2007; 42: 1391–1403Published online in Wiley InterScience(www.interscience.wiley.com) DOI: 10.1002/jms.1314

SPECIAL FEATURE:PERSPECTIVE

High-sensitivity analysis of specific peptides incomplex samples by selected MS/MS ion monitoringand linear ion trap mass spectrometry: Application tobiological studies

Inmaculada Jorge,1 Elisabet Miro Casas,2 Margarita Villar,1 Inmaculada Ortega-Perez,3

Daniel Lopez-Ferrer,1 Antonio Martınez-Ruiz,3 Monica Carrera,5 Anabel Marina,1

Pablo Martınez,1 Horacio Serrano,1,6 Benito Canas,7 Felipe Were,3 Jose Manuel Gallardo,5

Santiago Lamas,4 Juan Miguel Redondo,3 David Garcıa-Dorado2 and Jesus Vazquez1∗

1 Protein Chemistry and Proteomics Laboratory, Centro de Biologıa Molecular Severo Ochoa, CSIC, Madrid, Spain2 Servicio de Cardiologia, Hospital Vall d’Hebron, Barcelona, Spain3 Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain4 Centro de Investigaciones Biologicas, CSIC, Madrid, Spain5 Instituto de Investigaciones Marinas, CSIC, Vigo, Spain6 Universidad de Puerto Rico en Arecibo, Puerto Rico, Spain7 Analytical Chemistry Department, Universidad Complutense de Madrid, Spain

Received 9 July 2007; Accepted 22 August 2007

Mass spectrometry (MS) is a technique of paramount importance in Proteomics, and developments in thisfield have been possible owing to novel MS instrumentation, experimental strategies, and bioinformaticstools. Today it is possible to identify and determine relative expression levels of thousands of proteinsin a biological system by MS analysis of peptides produced by proteolytic digestion. In some situations,however, the precise characterization of a particular peptide species in a very complex peptide mixtureis needed. While single-fragment ion-based scanning modes such as selected ion reaction monitoring(SIRM) or consecutive reaction monitoring (CRM) may be highly sensitive, they do not produce MS/MSinformation and their actual specificity must be determined in advance, a prerequisite that is not usuallymet in a basic research context. In such cases, the MS detector may be programmed to perform continuousMS/MS spectra on the peptide ion of interest in order to obtain structural information. This selectedMS/MS ion monitoring (SMIM) mode has a number of advantages that are fully exploited by MS detectorsthat, like the linear ion trap, are characterized by high scanning speeds. In this work, we show someapplications of this technique in the context of biological studies. These results were obtained by selectingan appropriate combination of scans according to the purpose of each one of these research scenarios.They include highly specific identification of proteins present in low amounts, characterization andrelative quantification of post-translational modifications such as phosphorylation and S-nitrosylation andspecies-specific peptide identification. Copyright 2007 John Wiley & Sons, Ltd.

KEYWORDS: proteomics; tandem mass spectrometry; linear ion trap; ion monitoring; post-translational modifications;protein identification

INTRODUCTION

The term proteome was first coined to describe the setof proteins encoded by a particular genome.1 Proteomics

ŁCorrespondence to: Jesus Vazquez, Centro de Biologıa MolecularSevero Ochoa, Universidad Autonoma de Madrid, 28049Cantoblanco, Madrid, Spain. E-mail: [email protected]

is concerned with determining the structure, expression,localization, biochemical activity, interactions, and cellularroles of as many proteins as possible. There has been agreat progress owing to novel instrumentation, experimen-tal strategies, and bioinformatics tools.2 Two decades afterthe discovery of electrospray ionization (ESI) and matrix-assisted laser desorption ionization (MALDI), which, at last,allowed the gentle ionization of large biomolecules, mass

Copyright 2007 John Wiley & Sons, Ltd.

1392 I. Jorge et al.

spectrometry (MS) has become a powerful tool in proteinanalysis and the key technology in Proteomics.3 The develop-ment and increased use of MS are broadly impacting researchin biology and medicine.4 Recent advances illustrate the roleof MS-based Proteomics as an indispensable tool for molecu-lar and cellular biology and for the emerging field of systemsbiology.5 These include study of protein–protein interac-tions via isolation of affinity-based protein complexes on asmall and proteome-wide scale,6 – 8 mapping of numerousorganelles,9 and generation of quantitative protein profilesfrom diverse species.

The success of MS is driven by both innovative instru-mentation designs, especially those operating on the time-of-flight or ion-trapping principles, and large-scale biochemicalstrategies, which use MS for the detection and identifica-tion of proteins. The precise identification of thousands ofproteins from complex samples is now possible: any humanprotein can now be identified directly using the informa-tion contained in genome databases on the basis of minimaldata retrieved by MS. Similar to what took place before ingenomics, the increase in the automation of sample han-dling, analysis, and interpretation of results is generatinga great number of qualitative and quantitative proteomicdata.10 Entire protein complexes, signaling pathways, andwhole organelles are being characterized. Post-translationalmodifications remain difficult to analyze but efficient genericstrategies are presently emerging.11

The quantification of protein expression is also possibleby MS, frequently using the information provided by thefragmentation spectra taken in triple quadrupole or iontrap instruments. MS/MS is required because the coelutionof isobaric peptides is not uncommon when LC/MS isapplied to very complex peptide mixtures. Quantificationusing only Full Scan MS, although possible, suffers from lackof specificity, especially, for experiments with very complexmatrices like blood. MS/MS fragmentation usually providesunique characteristic fragment ions for a given peptide,which combined with the specific precursor m/z value andthe retention time, may be used for the selective monitoringand quantification of the peptides under study.

LC/MS data may be acquired using the diverse operatingscanning procedures attainable with most of the tandemmass spectrometers used in proteomics: Full Scan, SelectedIon Monitoring (SIM), Selected Ion Reaction Monitoring(SIRM), and Consecutive Reaction Monitoring (CRM). In theFull Scan analysis, analyzer voltages change continuously toallow the selection of ions with diverse m/z. Consequently,only a fraction of a given ion is detected. Full Scanchromatograms are usually represented as the Total IonCurrent (TIC) of a wide m/z range versus time in a plotgiving information about the m/z of the ions eluting fromthe column at any given retention time. TIC chromatogramsare much like UV traces except for the richness in informationprovided by MS when compared to UV. Information aboutthe m/z, and hence the mass from LC eluting compoundsis deduced from TIC plots. Nevertheless, it is important tokeep in mind that the intact mass of a compound is not aunique identifier. Using the mass spectrometer in the FullScan mode, specific compounds can be monitored and their

elution time and ion intensity represented in selective m/zplots.

However, a much more sensitive detection is possiblewhen the MS analyzers are concentrated to analyze onlyone ion of interest; this kind of scan is termed SIM. In aquadrupole, this is achieved by fixing the voltages so that thepassage of ions is restricted to those with a given m/z value.When ion traps are used, only ions in a very narrow rangeof m/z are trapped. As the mass range set is narrower, thespecificity of the SIM experiment increases. Plotting the TICintensity for this reduced range versus time constitutes theSIM chromatogram. Only compounds with a given m/z valueare detected and plotted, making the detection very selective.Visualized peaks in the SIM plot may be minor componentsin the Full Scan TIC chromatogram. Nevertheless, in the SIMplots there is no structural information about the selectedion, leaving some uncertainty about the identity of themonitored compound. This uncertainty may be critical whenSIM plots are taken from very complicated samples, sinceseveral peptides may produce ions with similar m/z underESI ionization, making selective detection unreliable. In anycase, higher sensitivity may be attained by the SIM scanmode than with the Full Scan experiment because the massspectrometer dwells for a much longer time over a smallermass range.

SIRM is the method used by the majority of scientists toperform highly specific and accurate peptide quantificationby MS. The SIRM experiment is accomplished by selectingthe m/z of the precursor ion for MS/MS fragmentationand specifically monitoring only one of the fragment ionsproduced. One could think of this operation as the SIM ofa fragment ion. SIRM experiments deliver the trace of aunique fragment ion that can be monitored and quantifiedin the midst of a very complex matrix. This characteristicmakes the SIRM plot most suitable for sensitive andspecific quantification. SIRM plots are very simple, usuallycontaining only a single peak. The combination of the m/zvalues from a given precursor ion and from one of thefragment ions produced by it, is known as a ‘transition’and is represented as (precursor mass ! fragment mass). Insome MS detectors, like ion traps, multiple transitions maybe programmed so that one of the daughter ions resultingfrom precursor fragmentation may be further fragmented,obtaining a MS3 spectrum. CRM can then be achievedby selectively monitoring one of the MS3 fragment ions.Specificity greatly increases at the expense of sensitivity,since a certain proportion of ions are lost in each of thefragmentation/trapping events. However, by adjusting thetrapping mass windows and the fragmentation energies agood compromise may usually be attained that allows ahighly specific identification of the compound of interest.

While SIRM and CRM are the most sensitive scanningmodes for peptide identification, they do not produceMS/MS or MSn spectra. Hence it is necessary that the peptidehas been previously identified and that the fragments usedfor monitoring have been checked to be specific for thesepeptide species. Hence, they are only applicable to peptidespreviously identified and the fragments used for monitoringmust be specific for the precursor peptides. However, when

Copyright 2007 John Wiley & Sons, Ltd. J. Mass Spectrom. 2007; 42: 1391–1403DOI: 10.1002/jms

Selected MS/MS ion Monitoring 1393

dealing with very complex samples, sometimes, it is difficultto find a specific fragment ion. In such cases, the MS/MSspectrum of the peptide under study is of paramountimportance to confirm the nature of the species identified,to detect alterations in the sequence, or to find the exact siteof a post-translational modification. Nevertheless, obtainingan interpretable MS/MS spectrum is not always possiblewhen the MS detector is programmed to fragment the mostintense species eluting from the column. Sample complexitymay be too high to allow the fragmentation of all thepeptide species eluting at a given retention time, or thepeptide concentration too low to be detected in the Full Scanspectrum as signal may be masked by background noise. Inthese circumstances, the MS detector may be programmed toperform continuous MS/MS scans on the selected precursoralong the gradient. This operating mode, to which we referto here as ‘selected MS/MS ion monitoring’ (SMIM) hasa number of advantages over other conventional scanningmodes: an increased sensitivity is possible by averaging alarge number of MS/MS scans over a narrow elution time and‘virtual’ SIM chromatogram traces for the different fragmentions can be obtained. These advantages are particularlyevident with MS detectors, like linear ion traps (LITs),which have a high scanning speed. In the last years, wehad successfully applied in our laboratory, this kind ofscanning mode for the study of peptide species in the contextof several research projects. Some representative examplesare presented in this work.

EXPERIMENTAL PROCEDURES

For connexin 43 (Cx43) identification, a protein extract fromcardiomyocyte membranes (100 µg), prepared as described,12

was redissolved in a 40-µl buffer consisting of 8% (w/v)SDS, 10% (v/v) glycerol, 25 mM Tris–HCl (pH 6.8), 5%(v/v) ˇ-mercaptoethanol, 0.01% (w/v) bromophenol blue,and was separated by SDS-PAGE (10% polyacrylamide gel,8 ð 8 cm). The gel was stained with Coomassie BrilliantBlue R-250 (Bio-Rad). The gel lanes were cut into fivepieces (containing approximately 20 µg of protein each),and the piece corresponding to the 37–50 kDa region wassubjected to digestion with porcine trypsin (Promega) at asubstrate/protease ratio of 20 : 1 (w/w), and the peptideswere extracted as previously described.13

Identification of phosphorylation sites of NFATc2 byp38 kinase was carried out as follows. As much as 50 ngof constitutively active MKK6 (SIGMA), 3 µg of GST-p38,and 10 µg of GST-HA-mouse NFATc2-(amino acid 4–385)14

which were kindly provided by Dr J. Han (The ScrippsResearch Institute) were incubated at 37 °C for 30 minin a kinase buffer containing 20 mM HEPES (pH 7.9),25 mM ˇ-glycerophosphate, 20 mM MgCl2, 0.1 mM sodiumorthovanadate, and 2 mM dithiothreitol. The reaction tookplace in the presence of 250 µM ATP. Assays were stoppedwith Laemmli sample buffer and loaded onto a 7% SDS-PAGE gel or alternatively frozen for ‘in solution’ digestion.

Characterization of Hsp90 S-nitrosylation sites in crudecell extracts was performed as follows: EA.hy926 cells weregrown and treated with S-nitroso-L-cysteine as described.15,16

For preparing cell protein extracts, confluent cells werescrapaed and resuspended in nondenaturing lysis solu-tion [(50 mM Tris–HCl (pH 7.4), 300 mM NaCl, 5 mM EDTA,0.1 mM neocuproine, and 1% Triton X-100 plus proteaseinhibitor cocktail)], incubated in ice for 15 min, and cen-trifuged at 10 000 g, for 15 min at 4 °C. Supernatants werecollected and protein-quantified with the Bradford reagent(Bio-Rad), and subjected to trypsin digestion ‘in solution’,as described17 or separated by SDS-PAGE using 7.5% gelin refrigeration, under darkness. Pure Hsp90 was includedin SDS-PAGE gel in separate lanes as a control of the elec-trophoretic mobility. ‘In-gel’ digestion was performed asdescribed,13 under darkness.

For species identification from crude fish extracts,sarcoplasmic protein extracts from white muscle of eachof species were prepared according to a previous work.18

The resulting complex protein mixture was subjected to an‘in solution’ digestion with trypsin (Promega, Madison, WI,USA) at 1 : 50 protease-to-protein ratio for at least 12 h at37 °C. Digests were individually desalted and concentratedusing in-tip reverse-phase resins (ZipTip C18, Millipore) anddirectly analyzed by HPLC-MS/MS.

MS analysis of peptides was performed by using aSurveyor HPLC system coupled to an LTQ LIT massspectrometer (Thermo-Fisher) as described previously,14,15

with minor modifications. Peptides were concentrated anddesalted on an RP precolumn (0.32 ð 30 mm, BioBasic-18, Thermo Electron) and on-line eluted on an analyticalRP column (0.18 ð 150 mm BioBasic-18, Thermo Fisher),operating at 2 µl/min and using a 170-min gradient from5% to 40% B [solvent A: 0.1% formic acid (v/v), solventB: 0.1% formic acid (v/v), and 80% acetonitrile (v/v)]. TheLIT was programmed to perform a different set of scans percycle according to each experiment, as described in the text.Normalized collision energy was set to 35% and a 3-Da masswindow was used to fragment selected parent ions. Proteinidentification in Rattus norvegicus-uniprot.fasta database wascarried out as described using SEQUEST,17 and statisticalanalysis and determination of error rates were performedafter doing a normal and inverted database search by usingthe Probability Ratio method.19

RESULTS AND DISCUSSIONCombination of scanning modes in the linear iontrapPeptide analysis in electrospray-based mass spectrometersare usually performed by HPLC separation on-line with MSdetection; the detector is programmed so that a suitablescanning mode is applied according to the purpose ofthe analysis. In the 3D ion traps, a popular combinationof scanning modes is the so-called ‘triple-play’ method(Scheme 1(A)). In this method a survey scan (or ‘Full Scan’)is firstly used to detect the presence of potential peptide ionsat low resolutions, and then one of the ions is automaticallyselected for further analysis. A higher resolution scan is thenperformed in a short mass range around the selected ion(‘Zoom Scan’) and then an MS/MS spectrum is triggeredin order to produce fragment information for peptideidentification or characterization. Automatic ion selection



from the Full Scan is performed on the most intense speciesaccording to a set of rules that are applied to avoid arepeated fragmentation of the same peptide or contaminantion (usually referred to as ‘dynamic exclusion’ (DE); seeScheme 1). Zoom Scans are performed to determine peptidemass of the precursor ion with a mass accuracy better thanthat obtained in the Full Scan, and when isotopic peaksare well separated (as is the usual case for peptides withthree or lower charges); it also serves to determine chargestate. This information is sometimes very useful for a betterinterpretation of the MS/MS spectra. Zoom Scan spectra inthe 3D trap are often omitted in order to obtain a highernumber of peptide fragmentations at the expense of massaccuracy; in such cases, a cycle of three consecutive MS/MSscans is usually performed (Scheme 1(B)).

The LIT has a considerably higher scanning speed thanthe 3D traps;20 hence, a considerably higher number of scanscan be programmed in each cycle. For instance, the LITallows performance of six consecutive Zoom Scan–MS/MScombinations per cycle (Scheme 1(C)). This method isparticularly useful for large-scale peptide quantificationby stable isotope labeling; Zoom Scan spectra are used toresolve peptide peaks from different isotopes, thus allowingrelative peptide quantification, and the subsequent MS/MSspectra are used for peptide identification. Quantificationand identification are thus afforded at the same time. Inprevious works, the practical utility of this method ofquantification applied to 16O/18O stable isotope labeling hasbeen demonstrated in our laboratory.21,22 When the highestperformance in peptide identification of very complexsamples is pursued, the LIT may be programmed to performa large number of consecutive MS/MS scans per cycle(Scheme 1(D)) by using dynamic exclusion-based programs;this method is particularly useful for the analysis of verycomplex peptide mixtures, such as those produced by thedigestion of entire proteomes.

In biological studies, the characterization of specificpeptide species is often pursued. Besides, the concentrationof the protein under study is frequently very low makingit impossible to detect any peptide from the protein abovebackground noise. In some occasions, the ion is maskedby adjacent ions having higher intensity, or the samplecomplexity is too high for this ion to be selected forfragmentation among a complex pool of coeluting peptideions. In such cases, when the m/z of the precursor ion isknown a priori, the MS/MS detector may be programmed totrigger a blind, continuous fragmentation over the peptideion along the RP gradient; in this work we refer to this kindof analysis as SMIM. Note that the SMIM concept, as definedhere, does not refer to a new kind of scanning mode, but toa conventional MS/MS scan on a fixed precursor ion that isselected a priori and performed along the entire LC run. Sincemost of the chemical noise due to electrospray ionizationis eliminated when performing MS/MS fragmentation,interpretable MS/MS spectra may be obtained even in caseswhere the precursor ion is not detected in the Full Scanspectrum. Besides, several MS/MS spectra may be averagedtogether along the peptide elution peak, thus increasingthe signal-to-noise ratio.23 Multiple peptides may also be

selected for monitoring, so that the detector performs ineach cycle, sequential MS/MS scans over a list of prefixedpeptide ions. The LIT is particularly suitable for this kind ofscanning modes due to its high scanning speed; in a recentwork, we demonstrated that the continuous monitoringof 30 different peptide species is possible, obtaining morethan three interpretable MS/MS spectra per cycle.24 Thiscombination of scanning modes is illustrated in Scheme 1(E),where the ‘S’ subindex denotes that the mass of the precursorion to be fragmented is selected a priori.

Mixed combinations of automatic fragmentations trig-gered according to ions detected in the Full Scan, togetherwith SMIM scans may also be programmed in the samescanning cycle. For instance, in Scheme 1(F) a survey scanis programmed followed by a dependent MS/MS scan thatserves to make a general identification of the peptides presentin the sample; this scan is useful to detect potential prob-lems that arise from sample preparation. The dependentscan is then followed by SMIM scans over two differentpeptide species; however, by repeating the scans more thanonce over the same peptide ion, the MS detector may bemade to spend more time scanning over one of the pre-cursors. This may be useful when the concentration of oneof the species is lower than that of the other and a moreextensive spectrum averaging is required for this speciein order to produce interpretable MS/MS spectra. Besides,multiple fragmentations over the same ions are also pos-sible; for instance, Scheme 1(G) shows a scanning modewhere two peptide species are continuously monitored byperforming an MS/MS spectrum followed by two MS3 spec-tra over two fragments selected a priori. Some examples ofthe application of these scanning modes to resolve a rangeof problems encountered in biology are presented in thefollowing sections.

Identification of connexin 43 in a crude preparationof membrane proteinsConnexin is a family of proteins of paramount biolog-ical importance whose most important and best knownfunction is to form transmembrane channels, known asconnexons or hemichannels, at the gap junctions.25 – 28 Gapjunctions are involved in propagation of electrical impulsein myocardium28,29 as well as chemical communication inmany other cell types and tissues.30 – 34 Within the heart,ventricular cardiomyocytes express almost exclusively a con-nexin isoform with a molecular weight of 43 kDa (Cx43),which is one of the most important connexin isoforms.35

Several studies have shown that Cx43 can participatein important cell process by mechanisms independentfrom gap junction-mediated cell-to-cell communication.36,37

Recent studies have shown that Cx43 can be located inintracellular membranes, as the nuclear membrane38 and,in particular, the inner mitochondrial membrane.39 Theexact localization, molecular interactions, regulation, mech-anisms of translocation, and functions of Cx43 in mem-branes outside gap junction areas are the subject of anintense research effort. So far, studies aimed to answerthese questions have been almost exclusively based onantibody-based techniques, including immunoelectrophore-sis, immunoprecitation, immunohistochemistry, and gold



MS/MSS2

Full Scan

MS/MSDE

MS/MSS1

6

(F)

Full Scan

Zoom Scan

MS/MSDE

Full Scan

MS/MSDE3

Full Scan

Zoom Scan

MS/MSDE6

Full Scan

MS/MSDE14

Full Scan

MS/MSS1

MS3S1 S2

MS3S1 S3

MS/MSS4

MS3S4 S5

MS3S4 S6

3

(A)

(C)

(E)

Full Scan

MS/MSS1

MS/MSS2

MS/MSS30

(G)(B)

(D)

Scheme 1. Some possible combinations of scanning modes in the linear ion trap.

labelled immuno-electron microscopy. In this context, theanalysis of Cx43 by direct techniques such as MS in puri-fied membrane fractions, multiprotein aggregates, and otherpreparations should be an invaluable research tool.

In a previous work, we analyzed the protein compositionof a preparation of rat membrane proteins by subjectingthe sample to SDS-PAGE, cutting the gel lane into piecesand performing an ‘in-gel’ tryptic digestion; the resultingpeptides were identified by HPLC-MS/MS using an LITMS as described.40 The MS/MS spectra were then searchedagainst a rat protein database. In this experiment, the peptideTYIISILFK from Cx43 was found to give a match in firstposition with one of the MS/MS spectra from a precursor ionat m/z 549.1; however, the SEQUEST scores (Xcorr D 1.913,DCn D 0.42) were not considered as statistically significantat the 5% FDR identification level. Besides, a close inspectionof the MS/MS spectrum was not conclusive, and it remainedunclear whether this peptide was correctly identified or not(not shown). Another Cx43 peptide gave even a worse matchwith MS/MS spectra over a precursor ion at m/z 514.3,yielding an Xcorr D 1.927 and a DCn D 0.087. Since no otherpeptides from Cx43 were detected in this preparation, theseresults did not allow to establish whether Cx43 was presentin the sample or not.

In order to confirm unambiguously the presence of Cx43in the preparation of membrane proteins, an identical fractionfrom the same gel was analyzed by SMIM programming theLIT as schematized in Scheme 1(G). An m/z 400–1600 FullScan was first performed in order to check for the presenceof digested peptides as well as peptide separation along thegradient. Then, one selected MS/MS ion monitoring wasprogrammed on the doubly-charged precursor ion at m/z549.1 and two additional MS3 spectra were programmedon ions at m/z 720.5 and 833.6, two of the most intensefragments in the MS/MS spectrum recorded in the previousexperiment. A group of scans with the same structurewas also programmed for ion at m/z 514.3, and the twoscan groups were repeated three times per cycle (seeScheme 1(G)).

The results of this experiment are shown in Fig. 1.The base peak chromatogram trace of the survey scansrevealed the presence of abundant ionic species as wellas a good peak resolution (Fig. 1(A)), suggesting that thedigestion was performed with good yield and there were nochromatographic problems. The chromatogram trace at m/z549.1 of the survey scans also revealed the presence of severalionic species having this mass (Fig. 1(B)); however, theMS/MS spectra of these precursors at the most intense peakswere completely different to the spectrum detected in theprevious experiment, indicating that none of these prominentspecies corresponded to the putative Cx43 peptide. Aftera careful inspection of all the MS/MS spectra obtainedalong the time course, an MS/MS spectrum resemblingthat obtained previously was found at a retention timearound 42 min (Fig. 1(F)). This MS/MS spectrum revealed aprominent fragment at m/z 833.6; a chromatogram trace wasthen drawn by plotting the intensity of this fragment versusthe retention time (Fig. 1(C)). This trace was completelydifferent to that obtained for the precursor peptide ionfrom the survey scan (compare with Fig. 1(B)), revealingthe presence of a prominent, single species at this retentiontime. Since this ion corresponded to the y00

3C fragment from

peptide TYIISILFK, we assumed that the peak in Fig. 1(B)was indicative of peptide elution at this retention time.Averaging of the MS/MS scans on precursor ion at m/z549.1 around the peak apex produced a fragment spectrumwhere the signal-to-noise ratio was improved (Fig. 1(G)).Most of the fragments detected in this spectrum couldbe assigned to y00- or b-ion series; however, some of thefragments could not be interpreted as arising from thispeptide (indicated by arrows in Fig. 1(G)). The spectrum wasthen filtered by subtracting adjacent MS/MS spectra fromthe same precursor (shown in Fig. 1(H) and (I)) in orderto minimize nonspecific background noise. The resultingspectrum was fully consistent with the peptide sequence andall the significant peaks in the spectrum could be assigned(Fig. 1(J)).



549.1RT: 42.46 min

549.1

50

100

50

100

RT: 42.03-42.54 min

50

100

Right BG; RT: 44.57-45.41 min549.1

50

100

Left BG; RT: 41.95 min549.1

(F)

(G)

(H)

(I)

Retention Time (min)

Inte

nsi

ty

500000

1500000

1000000

Full Scan(A)

100000

300000

200000

549.1

42.37

(B)

100

200

300833.6

42.28(C)549.1

10 20 30 40 50 600

5

10

15549.1

720.5

296.3

42.23(E)

70

5

10

15

549.1

833.6

556.4

42.39(D)20

Inte

nsity

y”7+y”6

+

833.59720.50a2

+

237.11

600

T Y I I S I L F K

y2y8 y7 y6 y5 y4 y3

b8b2 b3 b4 b5 b6 b7

b80+

933.49

y”8+

996.62

b70+

786.47

b7+

804.48

y”5+

607.47

b5+

578.28

y”102+

540.42

y”4+

520.36b4

+

491.36b4

0+

473.0

a4+

463.28

y”3+

407.33

b3+

378.26

y”2+

294.23

b2+

265.07

200 300 400 1000500 800700 900

10

20

30

40

50

60

70

Inte

nsity

(J)

549.1

(K)b6

I I S I L F K

y6 y5 y4 y3 y2

b3 b4 b5

b40+

409.3

b4+

427.3

y”2+

294.2

b30+

296.1

b3+

314.2

y”3+

407.3 y”4+

520.4

b50+

522.4b6

0+

669.4

798.7

b5+

b6+

687.4815.6

0

2

4

6

8

10

12

Inte

nsity

300 400 500 600 700 800

a4+

399.9a5

+

512.4

[b60-II]+

443.3[b6

0-I]+

556.4

[b6-I]+

574.4

y”5+

607.7

a6+

y”6+

659.4

720.5

549.1 833.6

200 300 400 500 600 700

m/z

0

2

4

6

8

Inte

nsity

702.5

y”5+

607.4

685.5

b5+

574.5

b50+

556.4

a5+

546.4

y”4+

520.4y”3

+

407.3

b4+

427.3

y”2+

294.2

b3+

314.1

b30+

296.3

y20+

276.2

a4+

399.3

a3+

286.4

b40+

409.2

y50+

589.3

[b50-II]+

330.2

[b50-I]+

443.3

[b5-I]+

461.4

y3y4

I S I L F K

y5 y2

b3 b4 b5

(L)

549.1 720.5

b7*+

b7

b6*+

b7+

b6+

540.4

b6

Figure 1. Identification of peptide TYIISILFK from connexin 43 in a preparation of membrane proteins by selected MSn ionmonitoring. A crude preparation of membrane proteins was separated by SDS-PAGE and the region around 43 kDa was cut andtrypsin-digested. The resulting peptides were analyzed by HPLC on-line with LIT- MS as described in the text. The left panels showthe chromatograms corresponding to the base peak signals of most intense precursor ions (A) and daughter ions originating fromthe continuous fragmentation of precursor ion at m/z 549.1 which corresponds with the doubly-charged ion of the peptide (B) as wellas the MS/MS and MS3 traces from selected peptide fragments (C)–(E). Panels (F) – (J) show the raw MS/MS spectrum from thesame ion at the peak apex (F), the averaged spectra of scans around the peak (G), the background MS/MS spectra just before(H) and after the peak (I), and the final averaged and background-corrected MS/MS spectrum (J). Two averaged, uncorrected MS3

spectra from the same ion are shown in (K) and (L). Fragments corresponding to the loss of internal amino acids from b ions in(K) and (L) are indicated with the same nomenclature used previously41.



The peptide identity was unambiguously confirmedby inspecting the two MS3 spectra, which were alsoaveraged around the peak in Fig. 1(C). As shown in Fig. 1(K)and (L), fragmentation of the two selected y00 ions wasvery rich and allowed the confirmation of their entiresubsequence. Interestingly, some intense fragments weredetected corresponding to the loss of internal amino acidsfrom b ions. Fragments of this kind are observed withsome frequency and their formation is thought to arisefrom cyclation of b ions, as previously demonstrated inour laboratory41; these ions are usually very intense, aproperty that is attributed to enhanced stability due tocyclation.41 Interpretation of the same results correspondingto the 514.0 ion failed to demonstrate that this ion wasproduced by sequence VAQTDGVNVEMHLK of Cx43,indicating that this method can also be used to detect falsepositive assignations. Finally, as shown in Fig. 1(D) and(E), a very clean trace of the presence of the Cx43 peptidecould be constructed by plotting the intensity of one of themost intense fragments detected in the MS3 spectra. Takentogether, the three traces shown in Fig. 1(C), (D), and (E) canbe used as a diagnostic signal characteristic of the presenceof this peptide and hence of Cx43, in the sample. When thistrace is detected, the final and unambiguous confirmationof the presence of Cx43 is achieved by checking that thecorresponding MS/MS and MS3 spectra are as shown inFig. 1(J)–(L) This SMIM experiment demonstrates for thefirst time that Cx43 can be unequivocally identified in crudeextracts of membrane preparations, and opens the way tothe study of this important protein at the proteomics level.

Characterization of in vitro phosphorylation sitesin NFATc2 by p38 kinaseCharacterization of phosphorylation sites in proteins isone of the most challenging tasks in modern proteomicsdue to the low stoichiometry of these modifications. TheNuclear Factor of Activated T cells (NFAT) family oftranscription factors regulates the transcription of manygenes that are active in different cell systems includingthe immune, cardiovascular, and nervous systems.42 Itcomprises of four members termed as NFATc1, c2, c3, andc4, whose activity is modulated by the calcium-activatedphosphatase calcineurin. In resting cells, NFAT is presentin the cytosol in a highly phosphorylated form. Whenintracellular calcium levels rise, NFAT N-terminal domainis rapidly dephosphorylated by calcineurin, leading toits nuclear translocation and enhancing its transcriptionalactivity. Once the calcium levels are restored, NFAT isexported to the cytoplasm42 – 44 in a process dependent onphosphorylation by different protein kinases including c-junNH2-terminal kinase (JNK), p38, CK1,GSK3, PKA, MEKK1,CK1, and DYRK.45 – 53 However, the exact identification ofthe modified residues has been difficult to document largelybecause of the technical difficulties, and to date it is stillunclear how the activity of NFAT can be orchestered in ascenario of more that 30 possible phosphorylation sites.

High sensitivity characterization of phosphorylation sitescan be performed by SMIM of prefixed ions known tobe potentially phosphorylated. We have recently used thisapproach to specifically detect a dynamic phosphorylation

site in NFATc2 by JNK in cell cultures.14,24 Phosphorylationsites may in theory be detected with higher sensitivityusing a triple-quadrupole MS detector when scanning isconcentrated on a single daughter ion characteristic ofphosphopeptides (single reaction monitoring or SRM) or bymonitoring neutral loss of phosphate group. However, thesescanning modes do not produce an MS/MS spectrum andhence no structural information is obtained. For this reason,these scans are usually followed by an MS/MS spectrumthat is only triggered when a precursor ion of certainintensity is detected. Hence, if the aim of the experiment is toobtain structural phosphopeptide information by MS/MS,better results should be expected by selected ion MS/MSmonitoring in an ion trap, since this latter approach combinesthe higher trapping efficiency of this detector for fragmentions after collision-induced dissociation (CID) with thepossibility of performing extensive spectrum averaging, asshown in the previous section. An example of the highlyspecific degree of structural information that can be attainedby SMIM of phosphopeptides is shown in Fig. 2. In orderto characterize phosphorylation sites of NFATc2 by p38kinase, a preparation of recombinant NFATc2 was subjectedto in vitro phosphorylation in the presence of the kinase andthe protein was digested with trypsin. Among other sites,two potential sites suspected to be phosphorylated by thiskinase are present in tryptic peptide TSPDPTPVSTAPSK. Toanalyze whether these sites are phosphorylated, the peptidesresultant from tryptic digestion of the phosphorylatedprotein were analyzed by the SMIM technique by a scanningmethod similar to that of Scheme 1(E). The scanning cyclecontained a Full Scan followed by 22 MS/MS scans on a setof ions corresponding to nonphosphorylated and suspectedphosphorylated peptides from NFATc2. Among these, oneMS/MS scan was programmed on the nonphosphorylatedform of peptide TSPDPTPVSTAPSK (m/z 692.8), to confirmthe presence of the nonphosphorylated form of the peptide,which was used as internal control, and other on the mono-phosphorylated form (m/z 732.8). To detect the presenceof the modified peptide, scans were averaged using asliding time window of about 1 min, and the resultingaveraged spectra were manually analyzed. Fragment spectrain good agreement with peptide sequence were detected ata retention time of 2 min; however, two different MS/MSspectra were detected at slightly different retention times(Fig. 2). At 2.04 min, the averaged MS/MS spectrum wasconsistent with a phosphorylation at the second Thr residue(Fig. 2(D)); while at 3.23 min the averaged fragment spectrumindicated a phosphorylation at the first Ser residue (Fig. 2(E)).In both cases a prominent y00

10 fragment was observed,probably due to the presence of a Pro residue in 5th position,which favored N-terminal peptide cleavage at this position.Luckily, this fragment contained the phosphorylated residuein one case, but not in the other; hence this fragment ionwas 80 Da higher in the first case (compare Fig. 2(D) and(E)). Plotting of the chromatogram traces of the respectivey00

10 fragments served to locate the exact retention timesof each one of the phosphorylated species (Fig. 2(B) and(C)). Interestingly, the trace peak intensity of the phospho-Thr peptide was considerably higher than that of the



T TS D P Pt V S A P S KIn

tens

ity

0

20

40

60

80

100 2.04

1064.5 (y”10+)

732.8

(B)

0.00

0.02

0.04

0.06

0.08

0.10

2.63

1363.6 (y”13+)

732.8

(A)

Retention Time (min)

0 5 10 150.0

1.0

2.0

3.0

4.0

5.0

3.23

984.5 (y”10+)

732.8

(C)

b6b5 b8 b10 b11 b12 b14b9b4

y”14y”

13y”12 y”

10 y”8 y”

7 y”5 y”

3 y”2y”

4y”6y”

9y”11

T s P D P T TP V S SA Pb13b6b5 b10 b11 b12b4b2

y”13y”

12 y”10 y”

7y”9

(D)

732.8

(E)

732.8

200 300 400 500 600 700 800 900 1000 1100 1200

b4+

481.2

b2+

269.1

b06

+

662.1

b132+

659.5

b122+

619.2

y”122+

599.0

y”132+

682.9

y”7+

689.3

y”9+

887.5

y”12+

1196.6

y”10+

984.5

∆b6+

581.4

b52+

289.2

∆b62+

291.1

∆b04

+

335.2∆b0

5+

463.3

b05

+

562.5

b102+

532.2

∆b112+

518.8Inte

nsity

0.00.20.40.60.81.01.21.41.61.82.02.22.42.6

200 300 400 500 600 700 800 900 1000 1100 1200

y”12

2+

639.0

y”10

2+

532.9y”

8+

786.4

y”3

+

331.2

b4+

401.1

y”9

+

967.5

b8+

875.3

Inte

nsity y”

12+

1276.5

∆y”13

+

1265.6

b12+

1231.6

y”11

+

1179.4

b11+

1134.3

∆y”11

+

1081.5

y”10

+

1064.4

∆b11+

1036.5

∆b10+

965.4

∆y”12

+

1178.6

∆y”10

+

966.5

b142+

723.8

∆b8+

777.4

∆b142+

674.9

∆y”14

2+

684.0

y”6

+

590.4

∆y”12

2+

589.8

y”2

+

234.2

y”8

2+

393.2

b62+

340.2

∆y”7

2+

339.2

y”4

+

402.1 ∆b6+

581.2

∆y”9

2+

435.3

y”5

+

503.3 b9+

962.2

∆b9+

864.4

∆y”9

+

869.5

y”13

2+

682.4

y”9

2+

484.0

b5+

498.1

90

100

0

10

20

30

40

50

60

70

80

m/z

P

K

Figure 2. Characterization of two sites of phosphorylation by p38 kinase in the same peptide of NFATc2. A preparation ofrecombinant NFATc2 was subjected to in vitro phosphorylation in the presence of p38 kinase and the protein was digested withtrypsin and analyzed by LIT-MS/MS as described in the Material and Methods section. A selected MS/MS monitoring wasprogrammed on the doubly-charged precursor ion corresponding to the mono-phosphorylated form of peptide TSPDPTPVSTAPSK(m/z 732.8). The MS/MS trace of fragments at m/z 1363.6, 1064.5 and 984.5, are shown in the left panels (A)–(C). Averaged MS/MSspectra around the retention times of 2.04 min (D) and 3.23 min (E) are shown in the right panels. These spectra correspond to twodifferent sites of phosphorylation in the same peptide.

phospho-Ser peptide. Although the direct comparison ofthese traces must be done with caution, since the position ofthe phosphate moiety at the sequence backbone may affectpeptide fragmentation, in this case the intensity differencesare so evident that a reasonable conclusion from theexperiment is that p38 mediates NFATc2 phosphorylationpreferentially at the Thr residue. This information is veryuseful to direct further experiments aimed to demonstratethe functional role of phosphorylation at this residue onNFATc2 regulation.

Characterization of S-nitrosylation sites in crudecell extractsThe formation of a thionitrite, R–S–N O in a cysteine thiol,a protein modification called S-nitrosylation or S-nitrosation is

increasingly being considered as a new paradigm of cell sig-naling due to its particular features as a nonenzymatic post-translational modification produced by nitric oxide (NO) andrelated reactive nitrogen species.54,55 One of the features thatmakes it so interesting for cell signaling is its easy reversibil-ity in cellular environments by several biochemical reactions.At the same time this is also one of the major limitations forits detection as it is very labile in many of the standard ana-lytical processes (including many of the processes of samplepreparation for peptide identification by MS) and furtherimprovements in analytical techniques are needed to get adeeper insight into the functional roles of this modificationsin (patho) physiological settings. Indeed, the modification isspecifically lost in standard MALDI ionization,56 probably



due to its characteristic absorption band around 350 nm. Aderivatization method has been used to identify proteinsand peptides that are S-nitrosylated by substituting thismodification by a more stable biotinylation.57 However, thisis an indirect method, which relies on a series of succes-sive chemical steps, and a compromise has to be reachedbetween sensitivity and specificity58 in order to avoid pos-sible artifacts. We have characterized the S-nitrosylation ofHsp90, a chaperone involved in the activation of one of theNO-producing enzymes, endothelial nitric oxide synthase(eNOS).15 This modification inhibits the capacity of Hsp90to activate eNOS and it can be the molecular basis of anegative feedback loop regulating NO production. UsingESI-MS/MS we characterized a cysteine residue in Hsp90that was S-nitrosylated in vitro and were able to describe bycomplementary methods the presence of the modification inendothelial cells treated with a nitrosothiol and, more sub-tly, with eNOS agonists.15 However, further insight into themolecular mechanisms underlying these events requires thedirect characterization of these S-nitrosylated sites in cellu-lar extracts and without using chemical derivatization steps.Some encouraging results in this regard were obtained byusing the SMIM approach and are described as follows.

An initial attempt to characterize S-nitrosylation eventsin crude, unseparated cell extracts were performed bytryptic digestion of the crude extracts followed by directSMIM analysis of the peptide digest, concentrating the MSdetector on the S-nitrosylated peptide of Hsp90 previouslycharacterized. For this purpose the scanning method ofScheme 1(F) was used. It included a Full Scan followedby one MS/MS scan, triggered on the most intense ionsdetected in the previous scan using dynamic exclusion,one MS/MS scan on the non-nitrosylated peptide, andsix MS/MS scans on the putatively nitrosylated peptide.The first dependent MS/MS scan served to identify themost abundant peptides produced by protein digestion;this information was useful to check digestion yield andalso sequence coverage (not shown); a higher numberof scans were programmed on the modified peptide toconcentrate the MS detector most of the time on thefragmentation of this specie. In these experiments we wereable to produce a good MS/MS spectrum that unequivocallydemonstrated the presence of the nonmodified peptide(not shown), suggesting that the method was sensitiveenough to characterize potential modifications of this specie.Unluckily, when the MS/MS spectra produced from thefragmentation of the S-nitrosylated peptide were analyzed,a prominent contaminant was detected having the sameprecursor mass and retention time (not shown) making thecharacterization of the S-nitrosylated peptide impossible.This potential interference was expected due to the extremelyhigh complexity of the peptide mixture obtained after trypticdigestion of a whole proteome.

Since the analysis of the MS/MS spectrum from thecontaminant suggested that it was produced by singlepeptide specie (not shown), we tried to separate thecontaminant protein by a single prefractionation step.For this end the crude cellular extract was subjected toSDS-PAGE separation and the gel region corresponding

to the electrophoretic mobility of Hsp90 was sliced, cutinto pieces and subjected to ‘in-gel’ tryptic digestion.The peptide digest was then analyzed by the SMIMtechnique using the same scanning method. Results fromthis experiment are presented in Fig. 3. The nonmodifiedpeptide was detected at a retention time of 48.5 min,and the averaged scan had a good signal-to-noise ratio,allowing the identification of this peptide (Fig. 3(D)). TheS-nitrosylated form of this peptide was detected at aretention time of 50 min, and the MS/MS spectra wasvery similar, except for the expected displacement ofthe series due to the modification (Fig. 3(E)), and alsoconfirmed unequivocally the presence of an S-nitrosylationin the second Cys residue at position 7. Plotting of thetrace chromatogram of fragment y00

12C produced very clear

peaks that were specific for both the control and themodified peptides (Fig. 3(B) and (C)). Since the NO- moietywas not expected to have a significant influence on thepattern of peptide fragmentation (a supposition supportedon the similarity of MS/MS spectra in Fig. 3(D) and (E))the differences in intensity of these peaks suggested thatonly 7.1% of the peptide was in S-nitrosylated form,illustrating how the SMIM approach can in some casesbe used to yield relevant quantitative information of theS-nitrosylation status of a protein. Note that S-nitrosylationis a rather labile modification and that in this calculationwe ignore the possibility that the modification is partiallylost during the preparation of the sample. In general,relative quantifications of modifications, in relation to thecontrol protein concentrations, can be potentially achievedby the SMIM approach; these experiments would require thedetector to spend the same time on the analysis of the controland modified peptides, and a previous knowledge of therelative intensity of the fragments, obtained by analyzingexact amounts of synthetic peptides. Note that we arealso assuming that ion intensity is proportional to peptideconcentration, but we do not presently know the actuallinear dynamic range of protein quantification that we areachieving with the SMIM approach.

Species-specific identification by selected MS/MSion monitoring in crude fish extractsConventional identification of unprocessed fish is done bythe examination of their morphological characteristics. How-ever, owing to the development of the fishing industry,marine products can be processed, making the appreciationof their external morphological features often impossible.To avoid cases of substitution of certain fish species byothers with less commercial value, a number of global reg-ulations have been implemented.59 As a consequence, theseregulations make the development of the necessary analyt-ical tools to make possible distinguishing between closelyrelated species highly recommendable. DNA-based60,61 andprotein identification methodologies18,62 have been usedwith authentication purposes, but today these techniquesare tedious and time-consuming. In this sense, the couplingbetween the separation power of liquid chromatographyand the ability of MS using the SMIM scanning mode,have provided a rapid and reliable method for the speciesidentification in complex mixtures.63 – 65



0

100

0

100

0

100

5030

Inte

nsity

Time (min)

L V T S P C C I V T S T Y G W T A N M E R

y3y4y5y6y8y9y10y11y12y13y14y19 y18

b18b17b16b6 b7 b9 b15

y17

b15

1166.5

1416.5

1181.0

1416.5

(A)

(B)

(C)

0

100

y91127.4 y17

1931.6

b171783.5

y121416.5

y131515.4

y141629.1

y101229.3

y111315.3

y8964.2

y6721.2

b9946.1

y4549.2

b6601.2

y5620.1

b7733.9

b181897.6

b151611.2

b161712.2

50

y3435.1

y2+18

1009.7

y2+19

1060.41166.5

(D)

y171961.5

m/z

0

100

1000600 1400 1800

y121416.5

b181928.5

b171813.5

b161742.4

b1611641.5

y141629.1

y131515.4

y111315.3

y101229.3

y2+17

981.3y2+

181025.5

y2+19

1075.5y7907.3

y8964.2

y6721.2

b7733.9

b6601.1

y5620.1

y4549.2

50 b9946.1

y3435.1

y91127.5

b8847.3

1181.0(E)

48.52

50.03

L V T S P C c I V T S T Y G W T A N M E R

y3y4y5y6y7y8y9y10y11y12y13y14y17y18y19

b18b17b16b6 b7 b8 b9

Figure 3. Characterization of Hsp90 S-nitrosylation in a crude protein extract from endothelial cells.- A crude protein extract fromEA.hy926 cells treated with S-nitroso-L-cysteine was subjected to SDS-PAGE and the gel was Coomassie-Blue-stained. The regionof the gel corresponding to the position of Hsp90 was cut and subjected to ‘in-gel’ digestion with trypsin. The peptides were thendirectly analyzed by LIT-MS/MS as described in the Material and Methods section. The scanning program corresponded to that inScheme 1(F). The selected MS/MS spectra were programmed on the doubly charged precursor ion corresponding to unmodifiedpeptide LVTSPCCIVTSTYGWTANMER (m/z 1166.5), as well as on the S-nitrosylated form of the same peptide (m/z 1181.0). Thebase peak profile corresponding to Full Scan spectra is shown in (A). The traces of fragment at m/z 1416.5 (y00

12C) from the control

and from the modified peptide are shown in (B) and (C), respectively. Averaged MS/MS spectra around the peaks in (B) and (C) areshown in (D) and (E). The absolute intensity of peaks from the nonmodified peptide form (D) and S-nitrosylated peptide form (E) were2.45E6 and 1.75E5, respectively. The S-nitrosylated Cys residue is indicated by a lowercase letter.

In this section, we present an example of the applicationof SMIM approach to the rapid and accurate differentiation oftwo closely-related hake species, which coexist in the samefishing ground (Cape hakes). For this end, crude proteinextracts from fish meat extracted from individuals belongingto these two hake species were directly subjected to trypsindigestion and the resulting peptide pool was analyzed bySMIM. In this experiment the scanning method was similarto that of Scheme 1(E), but only two precursor ions at m/z590.3 and 597.3 were selected for monitoring, correspondingto the doubly-charged ions from two peptides that have thesame sequence except for a single Asp to Glu substitution inposition 4 (Fig. 4). These were species-specific peptides fromthe protein parvalbumin, a heat-stable sarcoplasmic proteinthat we had sequenced in these two species in previousstudies.66 As shown in Fig. 4(B), when the averaged MS/MSspectra produced by SMIM monitoring of the first precursorion from Merluccius paradoxus extracts was analyzed, a veryclear MS/MS spectrum was obtained around a retentiontime of 56.2 min, which gave a perfect agreement with thepeptide sequence. Tracing of the y00

8C fragment at m/z 910.3

of this precursor ion produced a highly specific peptidepeak at the exact retention time (Fig. 4(A)). A similar resultwas obtained with the second precursor ion in the extractsfrom Merluccius capensis; in this sample the specific peptideeluted at a retention time of 56.46 min, and the averagedMS/MS also corresponded clearly with the peptide sequence(Fig. 4(D)). The MS/MS was very similar to that producedby the other peptide, and the trace of the corresponding y00

8C

fragment (m/z 924.3) also produced a highly specific peak(Fig. 4(C)).

In clear contrast, the MS/MS spectrum correspondingto the second peptide could not be detected in the samplefrom M. paradoxus. This is illustrated by a chromatogramtrace where no sign could be detected of the presence of thispeptide (Fig. 4(A)). The same negative result was obtainedwhen the MS/MS spectrum characteristic of the first peptidewas searched in the analysis of the sample from M. capensis(Fig. 4(C)).

These results illustrate how peptide fragment chro-matograms produced by specific precursor ions can beused as highly specific traces measuring the presence of



0 20 40 60 80 100 120

Time (min)

0

100

0

100R

elat

ive

Abu

ndan

ce56.17 590.3

Merluccius paradoxus

200 400 600 800 1000

m/z

0

100

Rel

ativ

e A

bund

ance

910.3

666.2

795.31066.4

1009.3519.2

448.2

377.2 505.6225.0

934.3732.0270.0 581.9 804.2

∆m = 115

I-G-V-D-E-F-A-A-M-V-Kb3 b4 b7 b8 b9 b10

y”3y”4y”5y”6y”7y”8y”9y”10

y”10

y”9

y”8

y”7

y”6

y”5

y”4y”3b3

b4

b7 b8 b9 b10

0 20 40 60 80 100 120

Time (min)

0

100

0

100

Rel

ativ

e A

bund

ance

56.46

Merluccius capensis

200 400 600 800 1000 1200m/z

0

100R

elat

ive

Abu

ndan

ce924.3

795.3

666.3

1080.3

519.21023.4448.2

225.1377.2 817.2746.2528.2 948.1

675.3270.0

∆m = 129

y”3y”4y”5y”6y”7y”9y”10 y”8

I-G-V-E-E-F-A-A-M-V-Kb3 b4 b5 b6 b7 b8 b9 b10

y”3

y”4y”5

y”6

y”7

y”8

y”10

y”9b3

b4

b5 b6

b7 b8 b9

b10

b2

b2 b5

b5

b6

b6

y”2

y”2

b2

b2

y”2

y”2

910.3(A)

(B)

(C)

(D)

910.3590.3

924.3597.3 24.3597.3

590.3 597.3

Figure 4. Discrimination of two closely-related hake species (Merluccius paradoxus and Merluccius capensis) by SMIM of twoparvalbumin species-specific tryptic peptides in crude fish extracts. The protein extracts were digested with trypsin ‘in solution’ anddirectly analyzed by HPLC-MS/MS. The IT detector was set to perform a continuous operation in the MS/MS mode, beingfragmented only the doubly-charged ions corresponding to specific peptides at m/z 590.3 and 597.3. The traces of fragment y00

8C

(m/z 910.3 and 924.3, respectively) as a function of retention time are plotted in the upper chromatograms. The MS/MS spectraobtained at each of the peaks are shown in the lower panels. The selected species-specific peptides corresponded to the samepeptide sequence differing in only one amino acid. Note that the relative abundance scale is the same in all the panels.

protein sequences belonging to particular hake species,and suggest that a broad range of species could be mon-itored simultaneously in a single LC-MS/MS experimentusing the SMIM approach. In cases where the number ofionic species that are monitored is small (two to four),this scanning method can also be performed by usinga conventional 3D ion trap; we have extensively usedthis approach using an LCQ Deca XP (Thermo Fisher)with good results regarding identification of species-specificpeptides.65,66 This approach can be easily automated andeven used to quantify the presence of small amounts ofcontaminating species in seafood products. The methodol-ogy provides a rapid, sensitive, and effective identificationtool that could be used by the authorities responsible forthe supervision of safety, quality, and labeling to guar-antee an appropriate composition of the foodstuffs to theconsumers and even to avoid problems of disloyal competi-tion.

CONCLUSIONS

In this paper, we report an array of results obtained from theapplication of the selected MS/MS ion monitoring techniqueto different biological problems. The SMIM approach can beperformed in the LIT using different combination of scansup to 30 scans per cycle. By combining free MS/MS scanswith Zoom Scans and selected MS/MS and MSn ion scansand adjusting the number of times each scan is performed,analysis of peptide species can be performed in a number ofresearch scenarios. These include unequivocal identificationof membrane proteins in crude membrane extracts and thecharacterization of relevant post-translational modificationssuch as phosphorylation and S-nitrosylation sites in vitroand in vivo. Finally, the technique is also particularly suitablefor performing a rapid and highly specific identification ofspecies in unseparated protein extracts.



AcknowledgementsThe authors wish to thank to Joaquın Abian for his helpful comments.This work was supported by grants BIO2003-01805, BIO2006-10085,GR/SAL/0141/2004 (CAM), CAM BIO/0194/2006, the Fondo deInvestigaciones Sanitarias (Ministerio de Sanidad y Consumo,Instituto Salud Carlos III, RECAVA), and by an institutional grantby Fundacion Ramon Areces to CBMSO.

REFERENCES1. Wilkins MR, Pasquali C, Appel RD, Ou K, Golaz O, Sanchez JC,

Yan JX, Gooley AA, Hughes G, Humphery-Smith I, Williams KL,Hochstrasser DF. From proteins to proteomes: large scale proteinidentification by two-dimensional electrophoresis and aminoacid analysis. Biotechnology (N Y) 1996; 14: 61.

2. de Hoog CL, Mann M. Proteomics. Annual Review of Genomicsand Human Genetics 2004; 5: 267.

3. Lane CS. Mass spectrometry-based proteomics in the lifesciences. Cell and Molecular Life Sciences 2005; 62: 848.

4. Aebersold R, Mann M. Mass spectrometry-based proteomics.Nature 2003; 422: 198.

5. Tyers M, Mann M. From genomics to proteomics. Nature 2003;422: 193.

6. Neubauer G, Gottschalk A, Fabrizio P, Seraphin B, Luhrmann R,Mann M. Identification of the proteins of the yeast U1 smallnuclear ribonucleoprotein complex by mass spectrometry.Proceedings of National Academy of Sciences of the United Statesof America 1997; 94: 385.

7. Neubauer G, King A, Rappsilber J, Calvio C, Watson M, Ajuh P,Sleeman J, Lamond A, Mann M. Mass spectrometry and EST-database searching allows characterization of the multi-proteinspliceosome complex. Nature Genetics 1998; 20: 46.

8. Rout MP, Aitchison JD, Suprapto A, Hjertaas K, Zhao Y,Chait BT. The yeast nuclear pore complex: composition,architecture, and transport mechanism. Journal of Cell Biology2000; 148: 635.

9. Andersen JS, Mann M. Organellar proteomics: turninginventories into insights. EMBO Reports 2006; 7: 874.

10. Mann M, Hendrickson RC, Pandey A. Analysis of proteins andproteomes by mass spectrometry. Annual Review of Biochemistry2001; 70: 437.

11. Reinders J, Lewandrowski U, Moebius J, Wagner Y, Sick-mann A. Challenges in mass spectrometry-based proteomics.Proteomics 2004; 4: 3686.

12. Ruiz-Meana M, Garcia-Dorado D, Hofstaetter B, Piper HM,Soler-Soler J. Propagation of cardiomyocyte hypercontractureby passage of Na(C) through gap junctions. Circulation Research1999; 85: 280.

13. Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-geldigestion for mass spectrometric characterization of proteinsand proteomes. Nature Protocol 2006; 1: 2856.

14. Ortega-Perez I, Cano E, Were F, Villar M, Vazquez J,Redondo JM. c-Jun N-terminal kinase (JNK) positively regulatesNFATc2 transactivation through phosphorylation within the N-terminal regulatory domain. Journal of Biological Chemistry 2005;280: 20867.

15. Martinez-Ruiz A, Villanueva L, Gonzalez de Orduna C,Lopez-Ferrer D, Higueras MA, Tarin C, Rodriguez-Crespo I,Vazquez J, Lamas S. S-nitrosylation of Hsp90 promotes theinhibition of its ATPase and endothelial nitric oxide synthaseregulatory activities. Proceedings of the National Academy ofSciences of the United States of America 2005; 102: 8525.

16. Martinez-Ruiz A, Lamas S. Detection and proteomicidentification of S-nitrosylated proteins in endothelial cells.Archives of Biochemistry and Biophysics 2004; 423: 192.

17. Lopez-Ferrer D, Martinez-Bartolome S, Villar M, Campillos M,Martin-Maroto F, Vazquez J. Statistical model for large-scalepeptide identification in databases from tandem mass spectrausing SEQUEST. Analytical Chemistry 2004; 76: 6853.

18. Carrera M, Canas B, Pineiro C, Vazquez J, Gallardo JM.Identification of commercial hake and grenadier species by

proteomic analysis of the parvalbumin fraction. Proteomics 2006;6: 5278.

19. Martınez-Bartolome S, Martın-Maroto F, Navarro PJ, Lopez-Ferrer D, Ramos-Fernandez A, Villar M, Garcıa-Ruiz JP,Vazquez J. Properties of average score distributions ofSEQUEST: the Probability ratio method. Molecular and CellularProteomics 2007; submitted.

20. Mayya V, Rezaul K, Cong YS, Han D. Systematic comparison ofa two-dimensional ion trap and a three-dimensional ion trapmass spectrometer in proteomics. Molecular and Cell Proteomics2005; 4: 214.

21. Lopez-Ferrer D, Ramos-Fernandez A, Martinez-Bartolome S,Garcia-Ruiz P, Vazquez J. Quantitative proteomics using(16)O/(18)O labeling and linear ion trap mass spectrometry.Proteomics 2006; 6 (Suppl. 1): S4.

22. Ramos-Fernandez A, Lopez-Ferrer D, Vazquez J. Improvedmethod for differential expression proteomics using trypsin-catalyzed 18O labeling with a correction for labeling efficiency.Molecular and Cell Proteomics 2007; 6: 1274.

23. Marina A, Garcia MA, Albar JP, Yague J, Lopez de Castro JA,Vazquez J. High-sensitivity analysis and sequencing of peptidesand proteins by quadrupole ion trap mass spectrometry. Journalof Mass Spectrometry 1999; 34: 17.

24. Villar M, Ortega-Perez I, Were F, Cano E, Redondo JM,Vazquez J. Systematic characterization of phosphorylation sitesin NFATc2 by linear ion trap mass spectrometry. Proteomics 2006;6 (Suppl. 1): S16.

25. Harris AL. Emerging issues of connexin channels: biophysicsfills the gap. Quarterly Reviews of Biophysics 2001; 34: 325.

26. Sosinsky GE, Nicholson BJ. Structural organization of gapjunction channels. Biochimica et Biophysica Acta-Biomembranes2005; 1711: 99.

27. Saez JC, Berthoud VM, Branes MC, Martinez AD, Beyer EC.Plasma membrane channels formed by connexins: theirregulation and functions. Physiological Reviews 2003; 83: 1359.

28. Sohl G, Willecke K. Gap junctions and the connexin proteinfamily. Cardiovascular Research 2004; 62: 228.

29. van Veen AA, van Rijen HV, Opthof T. Cardiac gap junctionchannels: modulation of expression and channel properties.Cardiovascular Research 2001; 51: 217.

30. Dermietzel R, Hertberg EL, Kessler JA, Spray DC. Gap junctionsbetween cultured astrocytes: immunocytochemical, molecular,and electrophysiological analysis. Journal of Neuroscience 1991;11: 1421.

31. Dermietzel R, Spray DC. Gap junctions in the brain: where, whattype, how many and why?. Trends in Neurosciences 1993; 16: 186.

32. Duval N, Gomes D, Calaora V, Calabrese A, Meda P,Bruzzone R. Cell coupling and Cx43 expression in embryonicmouse neural progenitor cells. Journal of Cell Science 2002; 115:3241.

33. Iino S, Asamoto K, Nojyo Y. Heterogeneous distribution of a gapjunction protein, connexin43, in the gastroduodenal junction ofthe guinea pig. Autonomic Neuroscience 2001; 93: 8.

34. Rozental R, Srinivas M, Gokhan S, Urban M, Dermietzel R,Kessler JA, Spray DC, Mehler MF. Temporal expression ofneuronal connexins during hippocampal ontogeny. BrainResearch Reviews 2000; 32: 57.

35. Severs NJ. The cardiac gap junction and intercalated disc.International Journal of Cardiology 1990; 26: 137.

36. Jiang JX, Gu S. Gap junction- and hemichannel-independentactions of connexins. Biochimica et Biophysica Acta-Biomembranes2005; 1711: 208.

37. Stout C, Goodenough DA, Paul DL. Connexins: functionswithout junctions. Current Opinion in Cell Biology 2004; 16: 507.

38. Dang X, Doble BW, Kardami E. The carboxy-tail of connexin-43localizes to the nucleus and inhibits cell growth. Molecular andCellular Biochemistry 2003; 242: 35.

39. Rodriguez-Sinovas A, Boengler K, Cabestrero A, Gres P,Morente M, Ruiz-Meana M, Konietzka I, Miro E, Totzeck A,Heusch G, Schulz R, Garcia-Dorado D. Translocation ofconnexin 43 to the inner mitochondrial membrane of



cardiomyocytes through the heat shock protein 90-dependentTOM pathway and its importance for cardioprotection.Circulation Research 2006; 99: 93.

40. Serrano H, Jorge I, Martinez-Acedo P, Navarro PJ, Perez-Hernandez D, Miro E, Garcıa-Dorado D, Vazquez J. Quantita-tives proteomics of mitochondrial membrane proteins by sodiumdodecyl sulphate polyacrylamide gel electrophoresis, 16O/18Ostable isotope labeling and linear ion trap mass spectrometry.Proteomica 2007; 0: 29.

41. Yague J, Paradela A, Ramos M, Ogueta S, Marina A, Barahona F,Lopez de Castro JA, Vazquez J. Peptide rearrangement duringquadrupole ion trap fragmentation: added complexity toMS/MS spectra. Analytical Chemistry 2003; 75: 1524.

42. Horsley V, Pavlath GK. NFAT: ubiquitous regulator of celldifferentiation and adaptation. Journal of Cell Biology 2002; 156:771.

43. Crabtree GR. Calcium, calcineurin, and the control oftranscription. Journal of Biological Chemistry 2001; 276: 2313.

44. Hogan PG, Chen L, Nardone J, Rao A. Transcriptionalregulation by calcium, calcineurin, and NFAT. Genes andDevelopment 2003; 17: 2205.

45. Chow CW, Rincon M, Cavanagh J, Dickens M, Davis RJ. Nuclearaccumulation of NFAT4 opposed by the JNK signal transductionpathway. Science 1997; 278: 1638.

46. Beals CR, Sheridan CM, Turck CW, Gardner P, Crabtree GR.Nuclear export of NF-ATc enhanced by glycogen synthasekinase-3. Science 1997; 275: 1930.

47. Zhu J, Shibasaki F, Price R, Guillemot JC, Yano T, Dotsch V,Wagner G, Ferrara P, McKeon F. Intramolecular masking ofnuclear import signal on NF-AT4 by casein kinase I and MEKK1.Cell 1998; 93: 851.

48. Graef IA, Mermelstein PG, Stankunas K, Neilson JR, Deis-seroth K, Tsien RW, Crabtree GR. L-type calcium channels andGSK-3 regulate the activity of NF-ATc4 in hippocampal neurons.Nature 1999; 401: 703.

49. Chow CW, Davis RJ. Integration of calcium and cyclic AMPsignaling pathways by 14-3-3. Molecular and Cellular Biology2000; 20: 702.

50. Porter CM, Havens MA, Clipstone NA. Identification of aminoacid residues and protein kinases involved in the regulationof NFATc subcellular localization. Journal of Biological Chemistry2000; 275: 3543.

51. Neal JW, Clipstone NA. Glycogen synthase kinase-3 inhibits theDNA binding activity of NFATc. Journal of Biological Chemistry2001; 276: 3666.

52. Sheridan CM, Heist EK, Beals CR, Crabtree GR, Gardner P.Protein kinase A negatively modulates the nuclear accumulationof NF-ATc1 by priming for subsequent phosphorylation byglycogen synthase kinase-3. Journal of Biological Chemistry 2002;277: 48664.

53. Gwack Y, Sharma S, Nardone J, Tanasa B, Iuga A, Srikanth S,Okamura H, Bolton D, Feske S, Hogan PG, Rao A. A

genome-wide Drosophila RNAi screen identifies DYRK-familykinases as regulators of NFAT. Nature 2006; 441: 646.

54. Martinez-Ruiz A, Lamas S. S-nitrosylation: a potential newparadigm in signal transduction. Cardiovascular Research 2004;62: 43.

55. Hess DT, Matsumoto A, Kim SO, Marshall HE, Stamler JS.Protein S-nitrosylation: purview and parameters. Nature ReviewsMolecular Cell Biology 2005; 6: 150.

56. Kaneko R, Wada Y. Decomposition of protein nitrosothiolsinmatrix-assisted laser desorption/ionization and electrosprayionization mass spectrometry. Journal of Mass Spectrometry 2003;38: 526.

57. Jaffrey SR, Erdjument-Bromage H, Ferris CD, Tempst P,Snyder SH. Protein S-nitrosylation: a physiological signal forneuronal nitric oxide. Nature Cell Biology 2001; 3: 193.

58. Martinez-Ruiz A, Lamas S. Proteomic identification of S-nitrosylated proteins in endothelial cells. Methods in MolecularBiology 2007; 357: 215.

59. Law 121/2004, 23rd January, Ministry of Agriculture, Fish, andFood of Spain. B.O.E. 2004; 31: 4864–4868.

60. Perez M, Vieites JM, Presa P. ITS1-rDNA-based methodologyto identify world-wide hake species of the Genus Merluccius.Journal of Agricultural and Food Chemistry 2005; 53: 5239.

61. Quinteiro J, Vidal R, Izquierdo M, Sotelo CG, Chapela MJ,Perez-Martin RI, Rehbein H, Hold GL, Russell VJ, Pryde SE,Rosa C, Santos AT, Rey-Mendez M. Identification of Hakespecies (Merluccius Genus) using sequencing and PCR-RFLPanalysis of mitochondrial DNA control region sequences. Journalof Agricultural and Food Chemistry 2001; 49: 5108.

62. Pineiro C, Vazquez J, Marina AI, Barros-Velazquez J, Gal-lardo JM. Characterization and partial sequencing of species-specific sarcoplasmic polypeptides from commercial hakespecies by mass spectrometry following two-dimensional elec-trophoresis. Electrophoresis 2001; 22: 1545.

63. Lopez JL, Marina A, Alvarez G, Vazquez J. Application ofproteomics for fast identification of species-specific peptidesfrom marine species. Proteomics 2002; 2: 1658.

64. Lo AA, Hu A, Ho YP. Identification of microbial mixtures byLC-selective proteotypic-peptide analysis(SPA). Journal of MassSpectrometry 2006; 41: 1049.

65. Carrera M, Canas B, Pineiro C, Vazquez J, Gallardo JM. De novomass spectrometry sequencing and characterization of species-specific peptides from nucleoside diphosphate kinase B (NDK B)for the classification of commercial fish species belonging to thefamily Merlucciidae. Journal of Proteome Research 2007; 6: 3070.

66. Carrera M, Gallardo JM, Pineiro C, Canas B, Lopez-Ferrer D,Vazquez J. Procedure for the identification of the commercialspecies from the Merlucciidae family, necessary elementsand applications. CSIC registration number 200603287, Patentapproval pending, 2006.


High-sensitivity analysis of specific peptides in complex samples by selected MS/MS ion monitoring...

Documents

Transcript of High-sensitivity analysis of specific peptides in complex samples by selected MS/MS ion monitoring...