Enhanced Sample Multiplexing for Nitrotyrosine …rena/pdfs/publication24.pdf · Enhanced Sample Multiplexing for Nitrotyrosine-Modiﬁed Proteins Using Combined Precursor Isotopic

Enhanced Sample Multiplexing for Nitrotyrosine-Modified ProteinsUsing Combined Precursor Isotopic Labeling and Isobaric TaggingRena ̃ A. S. Robinson* and Adam R. Evans

Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania 15260, United States

ABSTRACT: Current strategies for identification and quanti-fication of 3-nitrotyrosine (3NT) post-translationally modifiedproteins (PTM) generally rely on biotin/avidin enrichment.Quantitative approaches have been demonstrated whichemploy isotopic labeling or isobaric tagging in order toquantify differences in the relative abundances of 3NT-modified proteins in two or potentially eight samples,respectively. Here, we present a novel strategy which usescombined precursor isotopic labeling and isobaric tagging(cPILOT) to increase the multiplexing capability of quantify-ing 3NT-modified proteins to 12 or 16 samples usingcommercially available tandem mass tags (TMT) or isobaric tags for relative and absolute quantification (iTRAQ), respectively.This strategy employs “light” and “heavy” labeled acetyl groups to block both N-termini and lysine residues of tryptic peptides.Next, 3NT is reduced to 3-aminotyrosine (3AT) using sodium dithionite followed by derivatization of light and heavy labeled3AT-peptides with either TMT or iTRAQ multiplex reagents. We demonstrate the proof-of-principle utility of cPILOT with invitro nitrated bovine serum albumin (BSA) and mouse splenic proteins using TMT0, TMT6, and iTRAQ8 reagents and discusslimitations of the strategy.

Tyrosine nitration is a common post-translational mod-ification (PTM) to proteins that occurs in vivo especially

under conditions of oxidative and nitrosative stress. Thismodification is associated with a number of pathologies1−6

including neurodegenerative disorders7 such as Alzheimer’s8−11

and Parkinson’s diseases12,13 and cardiovascular disease.7,14

Reaction of NO• with superoxide anion (O2•−) forms

peroxynitrite (ONOO−), which incorporates a NO2 group tothe side chains of Tyr (or Trp, Phe, Cys, and Met)15 residuesleading to formation of 3-nitrotyrosine (3NT). Consequencesof 3NT modification to proteins are that it can alter proteinstructure, lead to aggregation, increase susceptibility todegradation, and cause reduced enzymatic activity.15,16 Addi-tionally, nitration of proteins may disrupt cell signaling eventsthat rely on phosphorylation of tyrosine residues.15,17,18

Proteomic methods to identify nitrated proteins have beenrecently reviewed.5,14,18−20 The most widely used method todetect 3NT-modified proteins is gel electrophoresis combinedwith Western blotting detection using an anti-3NT monoclonalor polyclonal antibody or immunoprecipitation of 3NT-containing proteins.19−21 Two-dimensional Western blottingrelies however on mass spectrometry (MS) identification ofprotein spots and often provides little or no information on thespecific site of 3NT modification. Matrix-assisted laserdesorption/ionization (MALDI)-MS,22,23 electrospray tandemmass spectrometry (MS/MS),22,24,25 and liquid chromatog-raphy coupled with MS/MS have also been employed toidentify nitrated residues in tryptic peptides.26−28

Recently, a number of groups have employed enrichmentstrategies to isolate, identify, and quantify 3NT-modified

peptides.11,15,29−34 These approaches rely on reduction of theNO2 group to an NH2 group in Tyr to form 3-aminotyrosine(3AT). A biotin tag is incorporated to 3AT, and using avidinaffinity chromatography, biotinylated peptides can be isolated.Quantitative methods for determining differences in the relativeabundances of 3NT-modified proteins have been reportedwhich rely on amine-reactive chemistry.35−37 Isotopic labelingmethods which allow precursor ion quantification have beendemonstrated with 2H-phenylisothiocyanate37 and 1-(6-methyl-[D0/D3]-nicotinoyloxy)succinimide.36 In a recent report,35 theisobaric tag for relative and absolute quantitation (iTRAQ)reagent was used to tag 3AT peptides (after the blocking of N-termini and lysine residues with an acetyl group). In thisapproach, no enrichment of 3NT-peptides is carried out;however, upon doing precursor ion scanning for specificiTRAQ reporter ions (e.g., m/z 114 and 117) 3NT-modifiedpeptides can be easily identified.35 Although it was onlypartially demonstrated, this approach can quantify a total ofeight samples using the iTRAQ 8-plex reagent kit.Herein, we sought to increase the multiplexing capability of

quantitative proteomics to assess differences in the relativeabundances of 3NT-modified proteins. Through combinedprecursor isotopic labeling and isobaric tagging (cPILOT), wedeveloped a novel strategy that uses commercially availabletandem mass tags (TMT) and iTRAQ reagents in order toincrease the number of samples that can be compared in a

Received: August 2, 2011Accepted: April 17, 2012Published: April 17, 2012

Article

pubs.acs.org/ac

© 2012 American Chemical Society 4677 dx.doi.org/10.1021/ac202000v | Anal. Chem. 2012, 84, 4677−4686

pubs.acs.org/ac

single analysis to 12 and 16, respectively. Additionally, thepresence of reporter ions in the mass spectra can be used toidentify nitrated peptides and help to eliminate false positiveidentifications that can arise from shotgun proteomics experi-ments.38 cPILOT builds upon currently used quantitativeproteomics strategies.35,39 Proof-of-principle cPILOT experi-ments are presented using simple mixtures of in vitro nitratedbovine serum albumin (BSA) peptides and peptides frommouse splenic proteins analyzed with nanoflow liquidchromatography tandem mass spectrometry (LC-MS/MS)using a LTQ-Orbitrap Velos MS.

■ METHODSIn Vitro Nitration of BSA. One milligram of BSA (Sigma

Aldrich; St. Louis, MO) was dissolved in 500 μL of phosphatebuffer saline (PBS), and a 50 μL aliquot of ∼40 mM ONOO−

(Millipore; Billerica, MA) was added to the solution. Thesample was incubated at room temperature for 30 min underconstant stirring (hereafter referred to as 3NT-BSA). Thereaction was quenched by flash freezing in liquid nitrogen. Acontrol BSA sample was treated in the same manner with theexception of addition of ONOO− to the buffer solution.Samples were stored at −80 °C until further use.Spleen Homogenization. Mouse spleen tissue was

homogenized in an ice-cold phosphate buffer saline (PBS)solution containing 8 M urea with 100 passes of a Wheatonhomogenizer. Homogenate solution was collected, sonicated,and centrifuged at 13 000 rpm for 10 min (4 °C). Supernatantswere collected and protein concentrations determined using theBCA assay according to the manufacturer’s instructions (PierceThermo; Rockford, IL). Samples were stored at −80 °C untilfurther use.Immunochemical Measurement of 3NT-BSA Level.

Control and 3NT-BSA samples (5 μL) were incubated at roomtemperature for 20 min with 5 μL of SDS (12% v/v) and 10 μLof Laemmli buffer [0.125 M Tryzma base, pH 6.8, SDS (4% v/v), and glycerol (20% v/v)]. Samples (250 ng) were blottedonto a nitrocellulose membrane and blocked with BSA (3% w/v) in a PBS solution containing 0.04% (v/v) Tween-20 and0.10 M NaCl (hereafter referred to as Wash blot). A 1:2000dilution of rabbit polyclonal anti-3NT primary antibody(Sigma-Aldrich) was added, and the blot was incubated for 3h on a rocker. The blot was rinsed for 5 min in Wash blot threetimes and subsequently incubated with a 1:8000 dilution ofantirabbit IgG alkaline phosphatase secondary antibody (Sigma-Aldrich) in Wash blot for 1 h at room temperature on a rocker.The membrane was washed as described above and developedusing a solution of 0.2 mM nitrotetrazolium blue chloride and0.4 mM 5-bromo-4-chloro-3-indolyl phosphate dipotassium inan alkaline phosphatase buffer (0.1 M Tris, 0.1 M NaCl, 5 mMMgCl2; pH 9.5). The dried blot was scanned on a Canonscanner and saved as a .TIF image.Trypsin Digestion. Desired amounts of samples were

subject to trypsin digestion as follows. DTT (0.25 M; ThermoFisher; Pittsburgh, PA) was added in a 1:40 protein:reagentmole excess and incubated at 37 °C for 2 h. Iodoacetamide(0.25 M; Acros Organics; Morris Plains, NJ) was added in a1:80 protein:reagent mole excess at 0 °C for 2 h in the darkfollowed by addition of 1:40 protein:reagent mole excess L-cysteine at room temperature for 30 min. Tris buffer (0.2 Mtris, 10 mM CaCl2, pH 8.0) was added to the mixture until thefinal volume was four times the original volume. TPCK-treatedtrypsin (Sigma Aldrich) was added to each sample at 2% w:w

enzyme:protein and incubated at 37 °C for 24 h. Samples wereflash frozen with liquid nitrogen and stored at −80 °C untilfurther use.

Synthesis of N-Acetoxy-H3-succinimide and N-Acetox-y-2H3-succinimide Reagents. The procedure for synthesis ofN-acetoxy-H3-succinimide and N-acetoxy-2H3-succinimide isdescribed elsewhere.39,40Briefly, 1.9478 g of N-hydroxysuccini-mide (NHS, Sigma Aldrich) was added to 4.8 mL of >99%acetic anhydride (Sigma Aldrich) or 1.9451 g of NHS wasadded to 4.4 mL of 99% 2H6-acetic anhydride (Sigma Aldrich).Reactions occurred at room temperature under nitrogen withconstant stirring for 15 h. White crystals were collected, washedthoroughly with hexane, and allowed to dry under vacuum.Product yield was 1.9909 and 1.5085 g for N-acetoxy-H3-succinimide and N-acetoxy-2H3-succinimide, respectively. NMRanalysis confirmed formation of the appropriate succinimideester (>95% purity).

3NT Reduction and Isobaric Tagging of Peptides. 3NTwas reduced to 3AT by the addition of 0.25 M sodiumdithionite in a 150 mol excess to peptide mixtures for a 30minute incubation at room temperature. For experiments usingthe TMT0 reagent, 50 μg of the light and heavy labeled 3AT-BSA peptides was pooled into a single mixture. To this sample(100 μg) one vial of TMT0 dissolved in anhydrous dimethylsulfoxide was added directly to the sample according to themanufacturer’s instructions. Reactions incubated at roomtemperature under constant stirring for 1 h. Excess hydroxyl-amine hydrochloride (Sigma Aldrich) was used to quench thereaction. For experiments using the TMT 6-plex reagents(hereafter referred to as TMT6), light and heavy labeled 3AT-BSA peptide samples were split into six mixtures each. Eachsample was then reacted with the appropriate TMT6 reagentaccording to the manufacturer’s instructions. Next, portions ofthe light and heavy labeled TMT6-derivatized samples weremixed into a new sample tube to generate the desired massratios for reporter ions (discussed below). Overall, the totalmass of light and heavy labeled peptides was equal afterpooling. For experiments using the iTRAQ 8-plex reagents(hereafter referred to as iTRAQ8), light and heavy labeledpeptide samples (100 μg) were each split into two equalaliquots. The appropriate iTRAQ8 reagent was dissolved in 70μL of ethanol and directly added to the light or heavy labeledpeptide samples. Reactions incubated at room temperatureunder constant stirring for 1 h. Light and heavy labeled peptidesamples were then pooled into a single mixture to generate thedesired iTRAQ8 reagent ratios (discussed below). Finally,samples were cleaned using Waters Oasis HLB C18 cartridgesand dried using a speedy-vac.

LC-MS/MS. Samples were separated on a nanoflow Eskigent2D LC system using reversed-phase chromatography. Sampleswere reconstituted in a water solution with 0.1% formic acid (1μg·μL−1) and separated on a nanoflow Eksigent 2D LC system.Buffer A was composed of 97:3 H2O:acetonitrile (ACN) with0.1% formic acid. Buffer B was composed of ACN with 0.1%formic acid. Samples were injected onto a trapping column(100 μm i.d. × 2 cm) packed with 200 Å C18 material(Michrom Bioresources Inc., Auburn, CA) using an autosam-pler. After a short wash, samples were eluted to a pulled-tipanalytical column (75 μm i.d. × 13.2 cm) packed with 100 ÅC18 material (Michrom Bioresources Inc.). Gradient elutionwas performed as follows (% buffer A:% buffer B): 90:10 for 10min, ramp to 70:30 over 90 min, ramp to 40:60 over 30 min,ramp to 20:80 over 5 min, hold for 10 min, followed by column

Analytical Chemistry Article

dx.doi.org/10.1021/ac202000v | Anal. Chem. 2012, 84, 4677−46864678

equilibration. Eluted peptides were ionized using the nanospraysource (∼1.7 kV) of an LTQ-Orbitrap Velos MS. The MS wasoperated using data-dependent acquisition with the followingparameters: full FT parent scans at 60 000 resolution over them/z range 300−1600, top seven most intense ions wereselected for CID (isolation width 3 m/z, 35% normalizedcollision energy) and HCD (10 ms activation time, isolationwidth 3 m/z, 45% normalized collision energy). CID and HCDMS/MS spectra were recorded in the LTQ and Orbitrap (7500resolution), respectively.Database Searching and Data Analysis. Raw data files

were searched against the BSA sequence (Swiss Prot EntryP02769) or mouse International Protein Index Database withSEQUEST in Proteome Discoverer 1.2 (Thermo). Databasesearching parameters include precursor mass tolerance of 15ppm, fragment mass tolerance of 1.0 Da, static modification ofcarbamidomethylation to cysteine, dynamic modifications oflight and heavy acetyl groups at the N-terminus and lysineresidues, methionine oxidation, and special modification forTMT- or iTRAQ-derivatized 3AT residues. Masses for thesemodifications were 239.163 (TMT0), 244.174 (TMT6), and319.216 Da (iTRAQ8). False discovery rates of p < 0.05 and0.01 were set using decoy database searching such that onlymedium and high confidence peptides, respectively, were usedfor analyses. Intensity and area information for light and heavylabeled peptides were provided in the software output (.msffile). Manual calculation of relative reporter ion peak intensity(or peak area) ratios was performed due to limitations in thesoftware output for this cPILOT strategy.

■ RESULTS AND DISCUSSION

Combined Precursor Isotopic Labeling and IsobaricTagging (cPILOT). Herein we present a novel approach,cPILOT, which combines isotopic labeling with isobaric taggingstrategies such as TMT and iTRAQ in order to extend thenumber of 3NT-modified samples that can be multiplexed in asingle experiment to 12 and 16, respectively. The strategyconsists of three primary steps as shown in Scheme 1:acetylation of primary amine groups, reduction of 3NT groupsto 3AT, and isobaric tagging (using TMT or iTRAQ) of 3ATgroups. Nitrated tryptic peptides are initially reacted with N-acetoxy-H3-succinimide ester or isotopically labeled N-acetox-y-2H3-succinimide ester in order to acetylate primary aminegroups (i.e., N-termini or lysine residues) with an isotopically

light (-H3) or heavy (-2H3) label, respectively. The second step

involves the use of sodium dithionite to reduce 3NT to 3AT.The third major step uses commercially available TMTreagents (Scheme 1) or the iTRAQ reagent to react with3AT according to the manufacturer’s instructions. Finally, thederivatized samples are mixed in a 1:1 light:heavy ratio andanalyzed by LC-MS/MS. We reduce sample loss and potentialloss of low-abundance 3NT-modified peptides by performing asingle cleanup step which occurs after isobaric tagging.

Validation of in Vitro Nitration and IsotopicAcetylation of BSA. Immunochemistry and LC-MS/MSexperiments were initially performed in order to confirmnitration and acetylation of BSA. BSA 3NT modification siteshave been well characterized using various enrichmentmethods.34,35,41−44 In total, BSA has 21 Tyr residues (Swiss

Scheme 1. Schematic Representation of the Chemical Derivatization Steps Performed in the cPILOT Strategy

Figure 1. Validation of in vitro nitration. (a) Immunochemical blot ofBSA incubated with buffer (control) and ONOO− after probing withprimary anti-3NT antibody. (b) Example MS/MS spectrum obtainedfor a doubly charged peak at tr = 117.4 min with m/z 829.3837 thathas been assigned as [(Acetyl)DAFLGSFLY(NO2)EYSR + 2H]2+.Corresponding CID generated b- and y-type fragment ions are labeledin the spectrum.



Prot Entry P02769) that could potentially be nitrated. Figure 1ashows an immunoblot with a dark band corresponding to the3NT-BSA sample. Under similar reaction conditions, withoutthe presence of ONOO−, BSA control samples showed noimmunoreactivity, verifying generation of in vitro nitrated BSA.Validation of the first step of cPILOT was tested by trypticallydigesting and acetylating (light label) 3NT-BSA followed byLC-MS/MS analysis on an LTQ-Orbitrap Velos MS. Nitratedpeptides should be shifted in mass according to the number ofadded acetyl (42.0106 Da) and nitro (44.9851) groups. Figure1b shows an example of the MS/MS spectrum obtained uponfragmentation of a peak with a precursor m/z 829.3837. On thebasis of the b- and y-type fragment ions detected in the CIDMS/MS analysis, the peptide has been assigned as [(Acetyl)-DAFLGSFLY(NO2)EYSR + 2H]2+. Acetylation of the N-terminus was observed in this spectrum as well as the y4 and y5singly charged fragment ions locating the position of nitrationto Tyr355. In this spectrum only Tyr355 was observed as 3NT;however, we also observe nitration of Tyr357 (data not shown).Overall, in these experiments we observed nitration of 15 Tyrresidues in which MS/MS spectra were manually validated. Thelocation of the 3NT-modification sites are as follows: Tyr54,Tyr108, Tyr161, Tyr163, Tyr171, Tyr173, Tyr179, Tyr180,Tyr355, Tyr357, Tyr364, Tyr376, Tyr424, Tyr475, and Tyr520.

3AT-BSA Isotopic Labeling and TMT0. Having validatedthe presence of nitration and acetylation, 3NT-BSA sampleswere reduced to 3AT-BSA, split into two equal aliquots, andisotopically labeled with either a light or a heavy acetyl group.Acetylated 3AT-BSA samples were then derivatized with theTMT0 reagent that generates reporter ion m/z 126 and pooledin a 1:1 ratio light:heavy mixture. Figure 2a shows an examplepair of doubly charged peaks at m/z 913.4019 and 916.4206that have a total mass shift of 6 Da. This mass shift indicates thepresence of two acetylations to the peptide. Upon independentCID MS/MS of the light and heavy precursor peaks (Figure2b) this pair has been identified as the peptide [(Acetyl)EY-(NH-TMT0)EATLEEccAK(Acetyl) + 2H]2+ . This trypticpeptide has been modified with both N-terminal and Lys acetylgroups, carbamidomethyl groups to Cys, and addition of theTMT0 reagent to Tyr376. Identification of this peptide isfurther supported by the HCD MS/MS spectra obtained forthe light and heavy precursor ions (Figure 2c). Confirmation ofthe TMT derivatization can be determined by the presence ofthe reporter ion m/z 126.1258 in the HCD spectra for both thelight and the heavy labeled precursor ions (Figure 2d).It should be noted that fragmentation of acetylated tryptic

peptides generates an immonium ion at m/z 143.118, whichupon loss of NH3 produces a fragment peak at m/z 126.092.

45

Figure 2. Demonstration of chemical derivatization using TMT0 reagents. (a) Integrated parent mass spectrum at tr = 54.00−54.45 min across them/z range 912−920 in which a doubly charged isotopic pair is observed at m/z 913.4019 and 916.4206 for the light and heavy labeled peptide,respectively. (b) Overlay of the MS/MS spectra obtained after CID of the light and heavy labeled peaks shown in a. It should be noted that eachpeptide peak was isolated individually. MS/MS spectra from the light and heavy labeled peptides have been superimposed in this spectrum. b- and y-type fragment ions (as labeled in the plot) have led to the peptide assignment of [(Acetyl)EY(NH-TMT0)EATLEEccAK(Acetyl) + 2H]2+. Thelowercase letter “c” in the peptide sequence corresponds to carbamidomethylated cysteine. (c) Overlay of the MS/MS spectra obtained after HCDof the light and heavy labeled peaks shown in a. (d) Zoomed-in low m/z region of the HCD spectra shown in c for the light (lower left) and heavy(upper right) labeled peptide peaks.



This immonium ion −NH3 fragment can be detected in theHCD MS/MS spectrum of the light precursor peak; however,at R = 7500 it is resolved from the TMT0 reporter ion m/z126.1258 in the Orbitrap MS. For lower resolution MS it willbe difficult to discern the contribution of the relative peakintensity that is derived from acetylated immonium ions −NH3and TMT reporter ions. This is not the case for the heavyprecursor ions, in which the immonium ion −NH3 is shifted by3 Da. For example, in the HCD MS/MS spectrum shown in theupper right of Figure 2d the immonium ion −NH3 occurs at m/z 129.1083. For TMT6 reagents, in which reporter ions aregenerated at m/z 126−131, lower resolution MS will not beable to distinguish immonium ion vs reporter ion signalgenerated at m/z 126 and 129. Therefore, these reporterswould have to be avoided in low-resolution instrumentsespecially for analysis of more complex mixtures in which therelative abundances across samples would not be known apriori. Another approach is to use iTRAQ reagents in whichreporter ions occur at lower m/z values (i.e., 113−121); thisapproach is presented below.Twelve Sample Multiplexing with TMT6 Reagents.

Isobaric tagging strategies have an inherent cost which isaugmented with the desired sample multiplexing capability foran experiment. This makes analyses expensive for projectsinvolving a large N, use of technical and/or biological replicates,and experimental validation (e.g., forward and reverse tagging).Herein, we demonstrate how cPILOT can be employed inorder to maximize the cost of isobaric tagging strategies for a

single experiment that involves identification of nitratedpeptides. Figure 3a shows a precursor ion MS spectrumobtained for a pair of peaks in 3AT-BSA whereby light andheavy acetylated peptides were derivatized with the TMT6

reagents (reporter ions m/z 126−131). The particular peptidepair shown at m/z 488.2824 and 489.7917 has been assigned to[(Acetyl)AW(NH-TMT6)SVAR + 2H]2+. The correspondingmass shift of 3 Da confirms the presence of only one acetylgroup which has been detected on the N-terminus. Addition-ally, this peptide contains a Trp residue that was nitrated invitro as validated by the mass shift corresponding to the TMT6

reagent in CID and HCD MS/MS spectra (data not shown).In order to assess the ability of this approach to measure

quantitative differences across multiple samples, we prepared a3NT-BSA sample and split it into 12 equal aliquots whereby 6aliquots were acetylated with the light label and an additional 6aliquots were acetylated with the heavy label. Each sample wasthen derivatized with a particular TMT6 reagent, and sampleswere pooled to generate the following mass ratios. For aliquotscontaining 3AT light labeled and TMT6 tagged peptides, the 6aliquots were mixed in a 1:50:10:1:2:0.5 ratio corresponding toreporter ions m/z 126:127:128:129:130:131. The 6 aliquotscontaining 3AT heavy labeled and TMT6 tagged peptides weremixed in a 1:0.5:2:1:10:50 ratio corresponding to reporter ionsm/z 126:127:128:129:130:131. Overall, the total mass ratios ofthe derivatized light labeled:heavy labeled tryptic peptides was1:1. This is reflected in the precursor peaks measured in theparent MS (Figure 3a). Figure 3b and 3c shows the presence of

Figure 3. Multiplexing of 12 independent samples using TMT6 reagents. (a) Integrated parent mass spectrum at tr = 32.65−33.99 min of the doublycharged isotopic pair observed at m/z 488.2824 and 489.7917 for the light and heavy labeled peptide, respectively, that has been assigned as[(Acetyl)AW(NH-TMT6)SVAR + 2H]2+. (b) Zoomed-in low m/z region of the HCD MS/MS spectra obtained upon isolation and fragmentation ofthe light labeled peak shown in a. (c) Zoomed-in low m/z region of the HCD MS/MS spectra obtained upon isolation and fragmentation of theheavy labeled peak shown in a.



each of the TMT6 reagents for the light and heavy labeledprecursor ions, respectively. The observed relative abundanceratios for the light precursor peak (Figure 3b) according topeak areas are 1.0:38:8.5:1.2:2.9:1.2, and the ratios for theheavy precursor peak (Figure 3c) are 1.0:5.6:3.6:1.7:4.0:11. Thecorrelation in measured versus expected reporter ion ratios islinear for both the light (R2 = 0.9991) and the heavy (R2 =0.7956) precursor ions. However, the measured reporter ionratios are compressed in some cases and overestimated inothers deviating from expected reporter ion ratios from samplepreparation. While some error in quantitation can be related toerrors during sample processing, upon closer inspection of theprecursor and CID MS/MS and HCD MS/MS spectra weattribute the inaccuracy in measured reporter ion ratios tooverlap of the isotopically labeled precursor ions that occursduring isolation in the LTQ. The [(Acetyl)AW(NH-TMT6)-SVAR + 2H]2+ peptide shown in Figure 3 is doubly chargedand contains a single label; thus, precursor ions are shifted inmass by 1.5 m/z. The isolation window used in these studieswas 3.0 m/z (±1.5 m/z). Therefore, it is highly possible thations from the heavy labeled precursor m/z were isolated withthe light labeled precursor m/z and vice versa. Isolation of bothprecursor ions in one MS scan event would cause an overlap in

the ion signal for each of the reporter ions and henceinaccuracies in the measured relative abundances. This isconsistent with the compressed and overestimated reporter ionratios that were detected in the [(Acetyl)AW(NH-TMT6)-SVAR + 2H]2+ and other peptides detected in this data set.Similar issues with underestimation of reporter ion abundanceratios have been previously discussed.46,47 Solutions to reducequantitation errors might involve the use of proton transferreactions to charge reduce precursor ions,48,49 incorporation ofa MS3 step,50 or use of different isotopically labeled reagents(e.g., dimethylation) that will increase the mass shift betweenlight and heavy labeled precursors in the parent mass scan.51

This is the first demonstration of a quantitative proteomicsapproach that couples precursor isotopic labeling with isobarictagging to enhance sample multiplexing capabilities for 3NT-modified peptides. Roughly speaking, the relative abundancesfor the cPILOT 12-plex study presented above are reflective ofthe expected abundances based on sample pooling for most ofthe reporter ions. The ability to distinguish immonium ions isindicated in these spectra (Figure 3b and 3c) by the tworesolved peaks at m/z 126 and 129 in both HCD MS/MSspectra. On the basis of experimental design, cPILOT offersinter- and intrasample quantification. For example, cPILOT

Figure 4. Demonstration of chemical derivatization using a single iTRAQ8 reagent (i.e., reporter ion m/z 113). (a) Integrated parent mass spectrumat tr = 44.93−47.83 min in which a doubly charged isotopic pair is observed at m/z 949.3803 and 952.3989 for the light and heavy labeled peptide,respectively, that has been assigned as [(Acetyl)ETY(NH-iTRAQ8)GDmADccEK(Acetyl) + 2H]2+. Lowercase “m” and “c” correspond tomethionine oxidation and cysteine carbamidomethylation, respectively. (b and c) Zoomed-in low m/z region of the HCD MS/MS spectra obtainedfor the light and heavy labeled peptide peaks shown in a, respectively.



could be employed to control (light) vs diseased (heavy)samples whereby information regarding the relative change inabundance across multiple N could be assessed at the level ofprecursor ion signal (i.e., average of all control or diseasebiological replicates), and intrasample variation can be assessedat the level of reporter ion signal.3AT-BSA Isotopic Labeling and iTRAQ. Figure 4 shows

the extension of cPILOT to iTRAQ multiplex reagents. Thedoubly charged peptide pair at m/z 949.3803 and 952.3989shifted in mass by 6 Da (Figure 4a) has been assigned to[(Acetyl)ETY(NH-iTRAQ8)GDmADccEK(Acetyl) + 2H]2+.The number of observed acetyl groups agrees with labeling ofthe N-terminus and a Lys residue, carbamidomethylation ofCys, and tagging of Tyr108 with the iTRAQ8 reagent thatproduces reporter ion m/z 113. Addition of the iTRAQ8

reagent to Tyr108 is also confirmed by the presence of thereporter ion m/z 113 in the HCD MS/MS spectra of both thelight and the heavy labeled precursor ions (Figure 4b and 4c,respectively). This peptide also contains an oxidized Metresidue in which oxidation could have occurred during the invitro nitration reaction. Due to the lower m/z range (i.e., 113−121) of iTRAQ reporter ions there is no overlap with theacetylated immonium ion −NH3 that occurs at m/z 126.Multiplexing with iTRAQ Reagents. We designed

another proof-of-principle experiment to assess whether ornot overlap in reporter ion abundances occurs during precursorion isolation. Nonoverlapping reporter ion m/z values forisotopically labeled peptides were purposefully selected for thisexperiment. Briefly, two aliquots of light labeled 3AT-BSA werederivatized with iTRAQ8 reagents that produce reporter ionsm/z 113 and 115 and mixed in a 1:1 ratio. Two equal aliquotsof heavy labeled 3AT-BSA were derivatized with iTRAQ8

reagents that produce reporter ions m/z 117 and 118 andmixed in a 1:4 ratio. Light and heavy labeled samples (1:1 ratio)were combined into a single mixture for analysis. Figure 5shows the low m/z region of the HCD MS/MS spectragenerated upon isolation and fragmentation of the light labeledprecursor peak at m/z 548.9827. On the basis of CID and HCD

MS/MS spectra (data not shown) this peak has been assignedas [(Acetyl)HPEY(NH-iTRAQ8)AVSVLLR + 3H]3+. The issueof overlapping precursor ion isolation for the isotopicallylabeled pairs is better demonstrated in this data than could bedelineated with the data presented in Figure 3. Theoretically,the HCD MS/MS spectrum (Figure 5) should only contain thereporter ions at m/z 113 and 115 since it was generated uponisolation and fragmentation of the light labeled precursor peak.However, there is a signal observed from the heavy labeledprecursor ion peak as the reporter ions at m/z 117 and 118 arealso present in the spectrum. Detection of the heavy labeledreporter ion peaks confirms that overlap in precursor ionisolation is occurring in the LTQ and is consistent with theinaccurate abundances measured in the 12-plex experiments(Figure 3b and 3c).Figure 5 only displays the HCD MS/MS spectrum obtained

for the light labeled precursor peak; however, accurate reporterion abundances for both isotopically labeled peaks areobserved. The expected reporter ion ratios for the light labeledpeak is 1:1 for m/z 113:115 and for the heavy labeled peak 1:4for m/z 117:118. For the light labeled precursor peak, themeasured reporter ion abundances based on peak intensity are1:1.1 for m/z 113:115. The heavy labeled precursor peakmeasured reporter ion abundances is 1:4.0 for m/z 117:118.Similar ratio values are obtained using peak areas as opposed topeak intensities for abundance calculations (see table in Figure5). These data provide evidence to support that cPILOT allowsdifferences in relative protein abundances to be measuredacross multiple samples.In order to demonstrate the potential capability of this

approach to detect nitrated peptides in complex mixtures, thecPILOT workflow was carried out on tryptic peptides frommouse splenic proteins. On the basis of the aforementionedissues, only a four-plex experiment was designed which allowedus to track the occurrence of precursor ion overlap and evaluatethe reporter ion quantitation. The peptide mixture was splitinto four individual aliquots, acetylated with either light orheavy labels, derivatized with iTRAQ8 reagents, and mixed in a1:1 ratio for m/z 113 and 114 and 1:1 ratio for m/z 115 and116, respectively. Figure 6a shows a precursor mass spectrumfor a doubly charged precursor peak pair at m/z 810.972 that isshifted by 3 Da. This pair has been assigned as the[(Acetyl)APGISY(NH2-iTRAQ

8)QRLVR + 2H]2+ peptidebelonging to phosphoinositide 3-kinase regulatory subunit 5protein (IPI00400180.1). From the HCD MS/MS spectraobtained from the light and heavy labeled precursors it isevident that this peptide was nitrated as indicated by thepresence of the four reporter ions. Additionally, the observedpeak intensity ratios for reporter ions are similar to theexpected values. It is still apparent from the MS/MS spectrathat there is overlap in the precursor isolation indicated by thepresence of the reporter ions at m/z 115 and 116 in the HCDspectra (Figure 6b) taken upon isolation of the light precursorpeak and the presence of m/z 113 and 114 in the HCD spectrataken upon isolation of the heavy precursor peak (Figure 6c).Under ideal conditions, only reporter ions at m/z 113 and 114should be present in HCD spectra from the light labeledprecursor (similarly only m/z 115 and 116 should be observedfor the heavy precursor). Overall, we identified 371 peptideswhich corresponds to 96 mouse splenic proteins. Five proteinswere identified that each contain a single nitrated peptide:phosphoinositide 3-kinase regulatory subunit 5 protein[(Acetyl)APGISY(NH2-iTRAQ

8)QRLVR], nebulin [(Acetyl)-

Figure 5. Quantitative demonstration of four multiplexed samplesusing iTRAQ8 reagents. Zoomed-in low m/z region of the HCD MS/MS spectra obtained upon isolation and fragmentation of the lightlabeled peak at tr = 67.48 min and m/z 548.9827 that has beenassigned as [(Acetyl)HPEY(NH-iTRAQ8)AVSVLLR + 3H]3+. Ob-served reporter ion peak intensities and peak areas are displayed in thetables (top).



DY(NH2-iTRAQ8)DLREDAISIK(Acetyl)], Ep300 protein

[(Acetyl)KVEGDM(Ox)Y(NH2-iTRAQ8)ESANNR], Isoform

1 of Bul lous pemphigo id ant igen 1 [(Acety l) -SIDELNSAWDSLNK(Acetyl)AW(NH2-iTRAQ

8)K(Acetyl)],and Npas3 protein [(Acetyl)VERY(NH2-iTRAQ

8)-VESEADLR]. We note that a low percentage (i.e., 1.6%) ofnitrated peptides was identified with modifications on Tyr orTrp residues from the entire data set. Additional enrichmentsteps (i.e., immunoprecipitation of 3NT-modified peptidesprior to cPILOT) may be necessary in order to identify largernumbers of nitrated peptides in complex mixtures.Caveats to obtaining accurate quantitative information with

cPILOT relies on minimal to no overlap in the precursor ionisolation for isotopically labeled peaks, efficient chemicalreactions, and minimal errors during sample preparation. TheN-acetoxy-2H3-succinimide reagent used to isotopically labeltryptic peptides in these studies can be substituted with otherreagents that will increase the mass shift between light andheavy labeled peptides, thus eliminating compression andoverestimation of reporter ion abundance ratios. Our laboratoryis currently exploring other precursor isotopic labeling reagentsand/or experimental schemes to improve upon these issues.Errors associated with sample loss were minimized during theseexperiments by performing “one-pot” reactions and using asingle sample cleanup step prior to LC-MS/MS analysis.

Finally, for application of cPILOT to complex peptide mixturesin which protein abundances are not known a priori it isnecessary to have efficient chemical reactions in order to reducecomplexity in data analysis and interpretation. Isotopic labelingand commercial TMT and iTRAQ reactions have efficienciesreported to be greater than 98%.15,52−54 Our results areconsistent with these reported efficiencies.

■ CONCLUSION

This work reports on the first quantitative proteomics approachthat combines precursor ion isolation with isobaric taggingstrategies such as TMT and iTRAQ for analysis of 3NT-modified proteins. cPILOT can extend sample multiplexingcapabilities to 12 and 16 samples with commercial TMT6 andiTRAQ8 reagents, respectively, and provide identification ofnitrated peptides through detection of reporter ions in HCDMS/MS spectra. Despite the observed limitations with thereagents used for this cPILOT strategy, accurate quantitativeinformation could be obtained with appropriate precursortagging reagents and/or improvements to experimental design.As 3NT modifications are low abundance, we also note that thisapproach can be combined with prior enrichment proceduresfor nitrated peptides such as immunoprecipitation or withselected ion monitoring experiments to search for reporter ions.Primary advantages with cPILOT include the ability to use

Figure 6. Application of cPILOT to mouse spleen tissue using iTRAQ8 reagents in a four-plex experiment. (a) Parent mass spectrum at tr = 105.03min of the doubly charged isotopic pair observed at m/z 810.972 and 812.481 for the light and heavy labeled peptide, respectively,[(Acetyl)APGISY(NH2-iTRAQ

8)QRLVR + 2H]2+ of the phosphoinositide 3-kinase regulatory subunit 5 protein (IPI00400180.1) protein fromspleen tissue. (b) Zoomed-in low m/z region of the HCD MS/MS spectra obtained upon isolation and fragmentation of the light labeled peakshown in a. (c) Zoomed-in low m/z region of the HCD MS/MS spectra obtained upon isolation and fragmentation of the heavy labeled peak shownin a.



reporter ion tags to identify nitrated peptides, some reductionof costs associated with purchasing commercial isobaric tag kits,and increasing the number of samples analyzed in a singleexperiment. The ability to increase multiplexing of proteomicsexperiments is an attractive area for many researchers that areexamining expression differences in numerous biologicalsamples and conditions.

■ AUTHOR INFORMATIONCorresponding Author*Phone: 412-624-8167. Fax: 412-624-8611. E-mail: [email protected].

NotesThe authors declare no competing financial interest.

■ ACKNOWLEDGMENTSThe authors acknowledge the University of Pittsburgh and theSociety of Analytical Chemists in Pittsburgh for funds tosupport this work. We also thank Sage Bowser in theDepartment of Chemistry NMR Facility for help withvalidation of synthesized reagents and Daret K. St. Clair atthe University of Kentucky for mouse spleen tissue.

■ REFERENCES(1) Greenacre, S. A.; Ischiropoulos, H. Free Radical Res. 2001, 34,541−81.(2) Ischiropoulos, H. Arch. Biochem. Biophys. 1998, 356, 1−11.(3) Ischiropoulos, H. Biochem. Biophys. Res. Commun. 2003, 305,776−83.(4) Ischiropoulos, H. Arch. Biochem. Biophys. 2009, 484, 117−21.(5) Ryberg, H.; Caidahl, K. J. Chromatogr. B: Anal. Technol. Biomed.Life Sci. 2007, 851, 160−71.(6) Yakovlev, V. A.; Mikkelsen, R. B. J. Recept. Signal TransductionRes. 2010, 30, 420−9.(7) Peluffo, G.; Radi, R. Cardiovasc. Res. 2007, 75, 291−302.(8) Butterfield, D. A.; Reed, T.; Sultana, R. Free Radical Res. 2011, 45,59−72.(9) Castegna, A.; Thongboonkerd, V.; Klein, J. B.; Lynn, B.;Markesbery, W. R.; Butterfield, D. A. J. Neurochem. 2003, 85, 1394−401.(10) Sultana, R.; Sowell, R. A.; Butterfield, D. A. In Oxidative NeuralInjury; Veasey , S. C., Ed.; Humana Press Inc.: New York, 2009; pp137−157.(11) Smith, M. A.; Richey Harris, P. L.; Sayre, L. M.; Beckman, J. S.;Perry, G. J. Neurosci. 1997, 17, 2653−7.(12) Chinta, S. J.; Andersen, J. K. Free Radical Res. 2011, 45, 53−8.(13) Tsang, A. H.; Chung, K. K. Biochim. Biophys. Acta 2009, 1792,643−50.(14) Lee, J. R.; Kim, J. K.; Lee, S. J.; Kim, K. P. Arch. Pharm. Res.2009, 32, 1109−18.(15) Abello, N.; Kerstjens, H. A.; Postma, D. S.; Bischoff, R. J.Proteome Res. 2009, 8, 3222−38.(16) Squier, T. C. Exp. Gerontol. 2001, 36, 1539−50.(17) Mallozzi, C.; Di Stasi, A. M.; Minetti, M. FEBS Lett. 2001, 503,189−95.(18) Monteiro, H. P.; Arai, R. J.; Travassos, L. R. Antioxid. RedoxSignal. 2008, 10, 843−89.(19) Bigelow, D. J.; Qian, W. J. Methods Enzymol. 2008, 440, 191−205.(20) MacMillan-Crow, L. A.; Cruthirds, D. L.; Ahki, K. M.; Sanders,P. W.; Thompson, J. A. Free Radical Biol. Med. 2001, 31, 1603−8.(21) Sultana, R.; Perluigi, M.; Robinson, R. A. S.; Reed, T.; Muharib,T.; Hughes, C. P.; Butterfield, D. A. Antioxid. Redox Signaling 2011,DOI: 10.1089/ars.2011.4109 .(22) Petersson, A. S.; Steen, H.; Kalume, D. E.; Caidahl, K.;Roepstorff, P. J. Mass Spectrom. 2001, 36, 616−25.

(23) Sarver, A.; Scheffler, N. K.; Shetlar, M. D.; Gibson, B. W. J. Am.Soc. Mass Spectrom. 2001, 12, 439−48.(24) Gokulrangan, G.; Zaidi, A.; Michaelis, M. L.; Schoneich, C. J.Neurochem. 2007, 100, 1494−504.(25) Kanski, J.; Hong, S. J.; Schoneich, C. J. Biol. Chem. 2005, 280,24261−6.(26) Jiao, K.; Mandapati, S.; Skipper, P. L.; Tannenbaum, S. R.;Wishnok, J. S. Anal. Biochem. 2001, 293, 43−52.(27) Sacksteder, C. A.; Qian, W. J.; Knyushko, T. V.; Wang, H.; Chin,M. H.; Lacan, G.; Melega, W. P.; Camp, D. G., II; Smith, R. D.; Smith,D. J.; Squier, T. C.; Bigelow, D. J. Biochemistry 2006, 45, 8009−22.(28) Willard, B. B.; Ruse, C. I.; Keightley, J. A.; Bond, M.; Kinter, M.Anal. Chem. 2003, 75, 2370−6.(29) Abello, N.; Barroso, B.; Kerstjens, H. A.; Postma, D. S.; Bischoff,R. Talanta 2010, 80, 1503−12.(30) Hsu, J. L.; Chen, S. H.; Li, D. T.; Shi, F. K. J. Proteome Res. 2007,6, 2376−83.(31) Nikov, G.; Bhat, V.; Wishnok, J. S.; Tannenbaum, S. R. Anal.Biochem. 2003, 320, 214−22.(32) Zhang, Q.; Qian, W. J.; Knyushko, T. V.; Clauss, T. R.; Purvine,S. O.; Moore, R. J.; Sacksteder, C. A.; Chin, M. H.; Smith, D. J.; Camp,D. G., II; Bigelow, D. J.; Smith, R. D. J. Proteome Res. 2007, 6, 2257−68.(33) Prokai-Tatrai, K.; Guo, J.; Prokai, L. Mol. Cell. Proteomics 2011,in press.(34) Lee, J. R.; Lee, S. J.; Kim, T. W.; Kim, J. K.; Park, H. S.; Kim, D.E.; Kim, K. P.; Yeo, W. S. Anal. Chem. 2009, 81, 6620−6.(35) Chiappetta, G.; Corbo, C.; Palmese, A.; Galli, F.; Piroddi, M.;Marino, G.; Amoresano, A. Proteomics 2009, 9, 1524−37.(36) Tsumoto, H.; Taguchi, R.; Kohda, K. Chem. Pharm. Bull.(Tokyo) 2010, 58, 488−94.(37) van Haandel, L.; Killmer, J.; Li, X.; Schoneich, C.; Stobaugh, J. F.Chromatographia 2008, 68, 507−516.(38) Stevens, S. M., Jr.; Prokai-Tatrai, K.; Prokai, L. Mol. Cell.Proteomics 2008, 7, 2442−51.(39) Chakraborty, A.; Regnier, F. E. J. Chromatogr. A 2002, 949,173−84.(40) Ji, J.; Chakraborty, A.; Geng, M.; Zhang, X.; Amini, A.; Bina, M.;Regnier, F. J. Chromatogr. B: Biomed. Sci. Appl. 2000, 745, 197−210.(41) Ghesquiere, B.; Colaert, N.; Helsens, K.; Dejager, L.; Vanhaute,C.; Verleysen, K.; Kas, K.; Timmerman, E.; Goethals, M.; Libert, C.;Vandekerckhove, J.; Gevaert, K. Mol. Cell. Proteomics 2009, 8, 2642−52.(42) Ghesquiere, B.; Goethals, M.; Van Damme, J.; Staes, A.;Timmerman, E.; Vandekerckhove, J.; Gevaert, K. Rapid Commun. MassSpectrom. 2006, 20, 2885−93.(43) Ghesquiere, B.; Helsens, K.; Vandekerckhove, J.; Gevaert, K.Proteomics 2011, 11, 1094−8.(44) Zhang, Y.; Yang, H.; Poschl, U. Anal. Bioanal. Chem. 2011, 399,459−71.(45) Kim, J. Y.; Kim, K. W.; Kwon, H. J.; Lee, D. W.; Yoo, J. S. Anal.Chem. 2002, 74, 5443−9.(46) Kuzyk, M. A.; Ohlund, L. B.; Elliott, M. H.; Smith, D.; Qian, H.;Delaney, A.; Hunter, C. L.; Borchers, C. H. Proteomics 2009, 9, 3328−3340.(47) Ow, S. Y.; Salim, M.; Noirel, J.; Evans, C.; Rehman, I.; Wright,P. C. J. Proteome Res. 2009, 8, 5347−55.(48) Vincent, C. E.; Ledvina, A. R.; Westphall, M. S.; Coon, J. J.Theutilization of ion/ion reactions for improved MS/MS spectralpurityProceedings of the 59th ASMS Conference on Mass Spectrometryand Allied Topics American Society of Mass Spectrometry: Denver,CO, 2011.(49) Wenger, C. D.; Lee, M. V.; Herbert, A. S.; McAlister, G. C.;Phanstiel, D. H.; Westphall, M. S.; Coon, J. J. Nat. Methods 2011, 8,933−935.(50) Ting, L.; Rad, R.; Gygi, S. P.; Haas, W. Nat. Methods 2011, 8,937−40.(51) Boersema, P. J.; Aye, T. T.; van Veen, T. A.; Heck, A. J.;Mohammed, S. Proteomics 2008, 8, 4624−32.



mailto:[email protected]

mailto:[email protected]

(52) Pichler, P.; Kocher, T.; Holzmann, J.; Mazanek, M.; Taus, T.;Ammerer, G.; Mechtler, K. Anal. Chem. 2010, 82, 6549−58.(53) Pierce, A.; Unwin, R. D.; Evans, C. A.; Griffiths, S.; Carney, L.;Zhang, L.; Jaworska, E.; Lee, C. F.; Blinco, D.; Okoniewski, M. J.;Miller, C. J.; Bitton, D. A.; Spooncer, E.; Whetton, A. D. Mol. Cell.Proteomics 2008, 7, 853−63.(54) Scholten, A.; Visser, N. F.; van den Heuvel, R. H.; Heck, A. J. J.Am. Soc. Mass Spectrom. 2006, 17, 983−94.



Documents

Enhanced Sample Multiplexing for Nitrotyrosine …rena/pdfs/publication24.pdf · Enhanced Sample Multiplexing for Nitrotyrosine-Modiﬁed Proteins Using Combined Precursor Isotopic