Global Differential Non-Gel Proteomics by Quantitative and Stable

18O-Labeling and Isolation of Amino Terminal Peptides Out of a Human Platelet Proteome. ..... 20204) of the Flanders Institute of Science and Technolo...
0 downloads 0 Views 97KB Size
Global Differential Non-Gel Proteomics by Quantitative and Stable Labeling of Tryptic Peptides with Oxygen-18 An Staes,† Hans Demol,† Jozef Van Damme, Lennart Martens, Joe1 l Vandekerckhove, and Kris Gevaert* Department of Medical Protein Research, Faculty of Medicine and Health Sciences, Ghent University, A. Baertsoenkaai 3, B-9000 Ghent, Belgium Received February 12, 2004

We describe a protocol for quantitative labeling of tryptic peptides with oxygen-18. Proteins are first digested in natural water with trypsin, the pH is then lowered to 4.5 and the mixture is dried. Oxygen18 water is added and two oxygen-18 atoms are incorporated at the peptides’ carboxyl termini. Trypsin is finally inactivated by cysteine alkylation under denaturing conditions, which blocks oxygen backexchange. The general value of this labeling strategy for differential proteomics is illustrated by the analysis and identification of several couples of differently labeled amino terminal peptides isolated from a human platelet proteome by a previously described chromatographic procedure. Keywords: nongel proteomics • mass spectrometry • oxygen-18 • differential analysis • COFRADIC

Introduction Until recently, mass spectrometry was not considered to be amenable to quantitative proteomics. This is mainly due to variable degrees of peptide detection depending upon the peptide’s nature (ease of ionization) and ionization suppression effects encountered when analyzing complex mixtures. However, mass spectrometry allows relative quantification of compounds by comparing ions that are identical though contain different stable isotopes discriminating them by a fixed mass difference. Due to the high sensitivity of mass spectrometry and the ease to operate it in a high-throughput mode, relative peptide quantification using one type of isotopic variants as references has become a widely spread method for differential proteomics.1-3 Stable heavy isotopes are either incorporated by metabolic labeling of growing cells or organisms in media enriched for the appropriate isotopes4-6 or alternatively, such isotopes are introduced by a chemical reaction with functional groups of amino acids.7-9 However, metabolic labeling of cells is expensive and can only be applied for cultured cells, affinity tags can only be applied to (a) specific subclass(es) of peptide(s) (e.g., cysteinyl peptides) and a chemical introduction of an isotope label is prone to losses of material. Techniques for analyzing the relative abundances of peptides based on trypsin-mediated incorporation of oxygen-18 isotopes at newly formed carboxyl termini have been reported.10-17 This enzymatic approach has also been used to label a peptide’s carboxylic end, thereby creating a recognizable signature for peptide fragments containing this group and thus helping in assigning the y-type of peptide fragments in spectra leading to more reliable protein identification.18-21 * To whom correspondence should be addressed. Tel: +32-92649358. Fax: +32-92649484. E-mail: [email protected]. † These authors contributed equally to this paper.

786

Journal of Proteome Research 2004, 3, 786-791

Published on Web 05/07/2004

Until recently, this labeling method was little used in quantitative studies of complex mixtures of peptides, probably because of variable degrees of labeling due to inefficient isotope incorporation and oxygen back-exchange. However, this tagging procedure, when quantitative, can be extremely useful for differential analysis since it has a number of advantages over metabolic or chemical tagging; (1) the reaction is simple and extra handling steps for the removal of excess reagents are not necessary, (2) oxygen-18 rich water is readily available and the volumes needed are economically justified, (3) every tryptic peptide (except those containing the proteins’ C-termini that do not end on a lysine or an arginine) is labeled with the same mass tag, (4) this mass difference allows sufficient separation between the isotope envelopes of the light and heavy peptide variants, while maintaining identical ionization behavior as well as highly similar chromatographic properties, and (5) this procedure is not biased toward the type of sample (e.g., peptides from digested proteomes from body fluids, tissues and biopsies can be labeled as efficiently as those from cultured cells). The evoked mass difference of 4 amu’s is most easily observed using MALDI-TOF mass spectrometers since in ESIbased mass spectrometers multiple charging of peptides occurs. The mass differences observed there are 2 and 1.3 amu’s for doubly and triply charged ions, respectively. However, ESIMS spectra can be de-convoluted and thereby singly charged ions are visualized, making it easier to observe the peptide couples and calculate their ratio. Given these important technical and physicochemical advantages, we here describe a study on trypsin-catalyzed oxygen-18 incorporation at the carboxyl end of tryptic peptides. We optimized conditions for complete incorporation of oxygen-18 isotopes as well as inhibition of back-exchange of the incorporated atoms with oxygen atoms from natural water. 10.1021/pr049956p CCC: $27.50

 2004 American Chemical Society

Differential Oxygen-18 Incorporation in Proteomics

We have recently developed a technology that is based upon a diagonal reversed-phase chromatography system and which allows tagless isolation of specific sets of peptides out of complex peptide mixtures (e.g., trypsin-digested proteomes). This technique is called COmbined FRActional DIagonal Chromatography (COFRADIC). One of the most interesting applications is the isolation of amino terminal peptides22 since this reduces the complexity of the final compound mixture to the highest degree; every protein is in theory represented by only one peptide (its amino terminal one). Furthermore, this approach allows us to monitor global processing of proteins, an important cellular phenomenon used to adjust protein function (e.g., apoptosis), drive protein trafficking (e.g., insertion of proteins in mitochondria) etc... We here provide a standard labeling protocol that can be applied to complex peptide mixtures, in this case derived from a total trypsin digest of human blood platelet proteins out of which amino terminal peptides were selected, thereby demonstrating that this type of labeling strategy might become a method-of-choice for future peptide-based differential proteome analyses.

Materials and Methods Materials. Horse heart myoglobin and tris(2-carboxyethyl)phosphine (TCEP) were from Sigma-Aldrich (St. Louis, MO). Sequencing-grade modified trypsin was from Promega Corporation (Madison, WI). 18O-rich water (93.7% pure) was purchased from ARC Laboratories (Amsterdam, The Netherlands). P10 ZipTips (U-C18) were from Millipore Corporation (Billerica, MA). Fresh human blood platelets were obtained from the Red Cross Blood Transfusion Centre Oost-Vlaanderen, Ghent, Belgium. Oxygen-18 Labeling Applied to a Tryptic Digest of Horse Heart Myoglobin. A 20-µg portion of horse heart myoglobin was digested overnight at 37 °C with 0.4 µg of trypsin (enzyme/ substrate ratio is 1/50, w/w) in 200 µL of 50 mM NH4HCO3 (pH 7.5). After digestion, aliquots of 15 µL (containing 90 pmol of the myoglobin digest) were transferred to new tubes. These mixtures were dried in a centrifugal vacuum concentrator after which 7.5 µL of 50 mM KH2PO4 (pH 4.5) was added. These mixtures were re-dried and redissolved in 30 µL of 18O-rich water (93.7% H218O). Incorporation of 18O-atoms proceeded overnight at 37 °C. Two stock solutions were prepared for the enzyme inactivation step; the first solution contained 1 mM of TCEP adjusted to pH 7 after addition of NaOH, whereas the second solution contained 1 mM of iodoacetamide in 3 M of guanidinium hydrochloride. A 10-µL portion of the TCEP stock solution was dried in one tube, and 20-µL of the iodoacetamide/guanidinium hydrochloride stock solution was dried in another tube. The labeled peptide mixture was transferred to the tube containing TCEP, and the reduction was allowed for 1 h at 37 °C after which this solution was transferred to the second tube containing the alkylating and denaturing agents and this mixture was put at 37 °C for 90 min. To monitor possible back-exchangesi.e., trypsin-based exchange of incorporated oxygen-18 isotopes with oxygen-16 isotopes when the samples are diluted in buffers containing natural watersthis peptide mixture containing the inactivated trypsin was diluted into the same volume of natural water (final concentration of H218O is about 50%) and incubated for several periods of time (1 h, 2 h, 4 h, 24 h, and 48 h) after which aliquots of 2 µL (corresponding to 3 pmol of digested myoglo-

research articles bin) were removed, desalted on a ZipTip according to the manufacturer’s procedure and spotted on a MALDI-target and analyzed in reflectron mode using a Bruker Ultraflex mass spectrometer. 18 O-Labeling and Isolation of Amino Terminal Peptides Out of a Human Platelet Proteome. The procedure to prepare a total proteome of human platelets and to digest with trypsin was essentially as described previously;22 briefly, isolated platelet proteins were alkylated on their cysteine residues using iodoacetamide and acetylated on their free amine groups. Subsequently, the proteins were digested with trypsin, which now only cleaves C-terminally to arginine residues. Prior to the primary RP-HPLC separation this proteome digest was split in two equal parts and one part was labeled with 18O-isotopes as described above while the second part went through the same procedure but now using natural water. Following inactivation of trypsin, the two peptide mixtures were re-mixed in a 1/1 ratio and fractionated a first time by RPHPLC in twelve primary fractions as described previously.22 Because of the employed COFRADIC sorting strategy, each primary fraction contains two types of peptides; internal ones with a free R-amino group and amino terminal ones with a blocked R-amino group. One such primary fraction was treated with TNBS (2,4,6-trinitrobenzenesulfonic acid) which introduces a very hydrophobic moiety on the free R-amines of the internal peptides. During the secondary, identical RP-HPLC separation, the TNBS-unaffected N-terminal peptides (since they already contain a blocked amino terminus) elute within the same time interval as during the primary separation whereas the modified internal peptides shift out of this interval and elute at much later times (hydrophobic shift). The amino terminal peptides present in one such a primary fraction were collected in a total of 8 sub-fractions (of 40 µL each) during one such a secondary separation. These subfractions were dried, redissolved in 20 µL of 0.1% formic acid in water and half of this sample was first analyzed by LC-MS on a Micromass Q-TOF mass spectrometer under the experimental conditions that have been described.22 The mass spectra were fed into Micromass’ proprietary MassLynx (version 3.5) software and were manually examined for the presence of peptide ion couples separated by 4 Da. The intensity of the monoisotopic ion of the light peptide (16O-tagged) was divided by the intensity of the monoisotopic ion of the heavy peptide (18O-tagged) and this value was then considered as the light/ heavy peptide ratio. Typically, only peptide ions with a signal/ noise ratio of more than three were considered for this analysis. The remaining half of the secondary fractions was then analyzed in LC-MS/MS mode using the same mass spectrometer and peak lists of the generated MS/MS spectra were used to identify the corresponding peptide using the MASCOT search algorithm (http://www.matrixscience.com). The following parameters were employed; enzyme setting was Arg-C, maximum number of missed cleavages was set to 1, fixed modifications were set to acetyl (K) and carbamidomethyl (C) and variable modifications were set to acetyl (N-term), deamidation (NQ), oxidation (M) and pyro-glu (N-term Q). Both the mass tolerances on the precursor and the fragment ions were set to 0.3 Da and the instrument setting was ESI-QUAD-TOF. A locally installed version of MASCOT was used and the human IPI database (http://www.ebi.ac.uk/IPI/IPIhelp.html) was searched. Only the spectra that were identified with a score above MASCOT’s identity threshold were considered positive. Journal of Proteome Research • Vol. 3, No. 4, 2004 787

research articles

Staes et al.

Results and Discussion Two Oxygen-18 Isotopes are Quantitatively and Stably Incorporated at the C-Terminal Carboxyl Group of Tryptic Peptides. The general concept of differential, nongel proteomics is to label one set of peptides with light isotopes (light peptides) and one set of peptides (isolated from a second proteome) with stable, heavy isotopes (heavy peptides). Following mass spectrometric analysis of the mixed peptide sets, the intensities of all observed ions (or only selected ones; e.g., the monoisotopic ions) belonging to either the light or the heavy peptides are summed and compared to each other; i.e., the light/heavy peptide ratio is determined. From these calculations, a distribution curve with its descriptive statistics (e.g., standard deviation) can be obtained. Subsequently, when a light/heavy peptide ratio differs from one in a statistically significant way (a deviation beyond the 95% confidence interval), this is indicative of a change in the abundance of a peptide in one proteome digest. Ultimately, when reproducible results are obtained, such peptides or proteins may be considered as ‘biomarkers’ which are differently represented in the mixture.23 In our approach, 16O and 18O are used as the light and heavy isotope, respectively. Oxygen incorporation takes place during trypsin-catalyzed hydrolysis of peptide bonds. When executed in H218O, this will lead to the C-terminal incorporation of two oxygen-18 atoms in every peptide formed by trypsin cleavage. Recently, it was noticed that a similar C-terminal specific oxygen-exchange can happen independent of a cleavage step at peptides terminating on arginine or lysine; the so-called post-cleavage trypsinmediated exchange, again leading to the addition of 4 Da when carried out in H218O.16,17 In most of these experiments, the mixture was immediately analyzed either by MALDI-MS or by direct injection in the ionization chamber of ESI mass spectrometers. In cases where the mixture was not directly analyzed but mixed with natural water during sample preparation, significant back-exchange with oxygen-16 was noticed. This back-exchange was also observed at pH values where trypsin hydrolysis is known to be inhibited. Given the fact that oxygen back exchange is a trypsincatalyzed reaction, taking place at pH values where the hydrolytic activity is quenched, we reasoned that any effort to denature trypsin would stop back exchange. When the trypsin/ peptide mixture was boiled for 5 min we noticed a slow (but significant) reactivation of oxygen exchange at room temperature. Incubation with 3 M guanidinium hydrochloride at various pH values blocked this activity as long as the denaturant was present, but was slowly restored after its removal. These results suggest partial re-naturation of the enzyme. Since the three-dimensional structure of trypsin is highly stabilized by disulfide bridges24 we used a combination of reduction followed by alkylation under denaturing conditions to block its activity. This experiment was carried out on a digest of myoglobin (Figure 1A) in which peptides were 18O-tagged according to the procedure described in the Materials and Methods section. Figure 1B shows that, even for prolonged periods of time (48 h) there is no back-exchange of incorporated 18O-atoms when trypsin is fully alkylated. As a control, the 18O-tagged peptides were diluted into the same volume of natural water (at pH of 4.5) with remaining noninactivated trypsin and even after a brief period of incubation (1 to 2 h), significant back-exchange of oxygen atoms was observed (Figure 1C). 788

Journal of Proteome Research • Vol. 3, No. 4, 2004

Figure 1. Stability of 18O-labeling. A MALDI-MS tryptic peptide mass map of horse myoglobin before 18O-labeling is shown in panel A. This peptide mixture was subsequently labeled by postcleavage incorporation of two 18O-atoms and diluted in the same volume of natural water at pH 4.5 with (panel B) or without (panel C) trypsin inactivation by reductive alkylation under denaturing conditions. After different periods of time, aliquots were removed from this diluted mixture and analyzed by MALDI MS. The MALDI MS spectra shown in panels B and C are derived from the peptide NH2-VEADIAGHGQEVLIR-COOH (the asterisk indicates the peptide ion that has incorporated two 18O-atoms) and illustrate that no oxygen exchange is observed when trypsin is inactivated even after prolonged periods of time (panel B). If trypsin is not inactivated, then rapid oxygen exchange is evident (panel C).

When we performed the isotope labeling at low pH-values (pH 2) we observed acid-catalyzed exchange even in the absence of trypsin and taking place at any carboxyl group present in the peptide. This was mentioned before by Stewart et al.12 and is illustrated here using a peptide containing a blocked C-terminal carboxyl group and free carboxyl groups on its aspartic and glutamic amino acids (Figure 2). Following overnight incubation at pH 2, a high number of peptides show varying degrees of 18O-incorporation, leading to a large number of isotopic variants. These experiments lead to the conclusion that posthydrolysis oxygen-labeling of tryptic peptides is optimal around pH 4. At this pH, the exchange rate is still as at higher pH-values (data not shown), while further unspecific cleavage or cleavage at slower sites is blocked. Such tryptic cleavages would lead to partial oxygen isotopic labeling and compromise differential

Differential Oxygen-18 Incorporation in Proteomics

Figure 2. Acid-catalyzed incorporation of 18O-atoms at aspartic and glutamic acids in peptides. The peptide Ac-VHHQKLVFFAEDVGSNK-CONH2 was incubated for 24 h in a citrate buffer at pH 2 in H218O (upper panel) and in natural water (lower panel). As evident from the wide isotope distribution, prolonged storage at low pH results in oxygen exchange on the side chains of acid amino acids (upper panel) and thus should be avoided in postcleavage 18O-labeling protocols.

analyses. Use of buffers with pH below pH 2 must also be avoided as they may trigger acid-catalyzed oxygen back exchange at all carboxyl functions present in the peptide. Buffers which may be used for prolonged storage of 18O-tagged peptides without any oxygen-exchange are 0.1% formic acid and 0.1% trifluoroacetic acid in water. Oxygen-18 Tagging as a Differential Quantitative Method in Global Proteomics. The procedure worked out above was used for a large scale differential proteomics study using total lysates of human blood platelets. To produce a less complex but still representative peptide mixture, we selected the N-terminal peptides of each constituent protein, discarding all internal peptides. Herefore, we used the COFRADIC (COmbined FRActional DIagonal Chromatography) procedure which was previously described in detail.22 Briefly, proteins were first alkylated at their cysteines and their free amino groups were blocked by acetylation. Trypsin cleavage now only occurs at arginine residues. The resulting peptide mixture is divided in two equal parts. One part was labeled with 18O, whereas the second part was passed through the same procedure but in natural water. The two peptide sets were remixed in a 1/1 ratio and out of this complex peptide mixture we isolated the N-terminal peptides while discarding all internal ones.22 One four min fraction (out of a total of twelve primary fractions), which was further taken through the TNBSmodification step and a consecutive secondary HPLC-separation was analyzed. The acetylated N-terminal peptides, eluting in the same time interval were collected in eight secondary fractions of 30 s each. Half of the material was used to measure the intensities of the peptide couples following LC-MS analysis of each of these 40 µL fractions. A total of 223 couples were observed and a histogram of their ratios is shown in Figure 3. Due to the fact that the peptides were often spread over two

research articles

Figure 3. Histogram and curve showing the distribution of the measured 16O/18O-ratio of N-terminal peptides isolated out of a human platelet proteome. The ratio values were obtained by dividing the intensity of the monoisotopic ion from the light peptide by that of the heavy peptide form. As evident from the histogram, most ratios center round the calculated average ratio of 1.10 with a 95% confidence interval between 0.79 and 1.53. A curve, showing the normal distribution of the measured ratios is superimposed on the histogram.

consecutive sub-fractions, the total number of unique peptide couples was 99. The remaining half was now used to perform a similar series of LC-runs now on-line coupled with fragmentation analysis. 31 peptide couples could be identified, corresponding to 26 different proteins (Table 1). Some proteins (e.g., R and β-fibrinogen and the pyruvate kinase 3 isoform 2) were identified by multiple peptides. As can be seen in Table 1 the measured ratios of such peptides only differ slightly from one another, indicating that one may eventually take an average of these ratios to more accurately determine the actual protein ratio. Extrapolation of these results for the peptides present in the twelve primary fractions, representing the complete proteome, would suggest a number of positively identified proteins close the previously reported 264 proteins.22 This illustrates that both differential 16O/18O peptide analysis and identification can go hand in hand without conversely being compromised by the increased complexities of the isotope envelopes. The base 2 logarithms of the measured peptide ratios showed a normal distribution from which the mean and the standard deviation were determined. These values were then used to back-calculate the average measured value as 1.10. This value may appear slightly higher than expected, yet this might be due to the fact that the 18O-rich water was not pure (only for 93.7%), leading to a bias toward the presence of light peptides. A 95% confidence interval for the measured ratios of peptide couples present in a 1/1 ratio is comprised between 0.79 and 1.53, which is consistent with values reported in other studies.25,26 Furthermore, this indicates that in future experiments, whenever the ratio of a peptide couple measured using the same experimental setup falls outside this interval, the concentration of this peptide differed significantly from 1/1 at the 95% confidence level between the two proteome digests, i.e. the peptide is significantly up- or down-regulated in one of the proteomes under study. Journal of Proteome Research • Vol. 3, No. 4, 2004 789

research articles

Staes et al.

Table 1. List of Peptides and Corresponding Proteins that Were Identified after Analysis of the Eight Secondary COFRADIC Fractionsa IPI-no.

1 IPI00003927 2 IPI00016638.1 3 IPI00026182 4 IPI00017614.6 5 IPI00295387.2 6 IPI00298497.3 7 IPI00293975.2 8 IPI00219018.1 9 IPI00013219 10 IPI00027444 11 IPI00027410 12 IPI00025252.1 13 IPI00031479.1 14 IPI00010796.1 15 IPI00003870.1

identified peptide

Ace-SHPSPQAKPSNPSNPR-COOH NH2-Q〈Pyr〉 KTGTAEMSSILEER-COOH Ace-ADLEEQLSDEEKVR-COOH Ace-AHYKAADSKR-COOH Ace-M〈Mox〉 ELEPELLLQEAR-COOH NH2-Q〈Pyr〉 GVNDNEEGFFSAR-COOH Ace-AAATQKKVER-COOH NH2-PLAGGEPVSLGSLR-COOH Ace-GKVKVGVNGFGR-COOH Ace-MDDIFTQCR-COOH Ace-M〈Mox〉 EQLSSANTR-COOH NH2-Q〈Pyr〉 HLGLVGGEEPPR-COOH Ace-SDVLELTDDNFESR-COOH Ace-AKVSSLIER-COOH Ace-DAPEEEDHVLVLR-COOH Ace-PLIPIVVEQTGR-COOH

19 IPI00010779.1 20 IPI00397834.1 21 IPI00021885.1

Ace-AAM〈Mox〉 ADTFLEHM〈Mox〉 CR-COOH Ace-AAM〈Mox〉 ADTFLEHMCR-COOH NH2-SLEVSPNPEPPEKPVR-COOH Ace-EGDFLAEGGGVR-COOH Ace-DSGEGDFLAEGGGVR-COOH Ace-ADSGEGDFLAEGGGVR-COOH NH2-NKDSHSLTTNIM〈Mox〉 EILR-COOH Ace-GLNSLEAVKR-COOH NH2-PQPPPDPLLLQR-COOH NH2-EYHTEKLVTSKGDKELR-COOH

22 IPI00021891 23 IPI00297779 24 IPI00027626.1 25 IPI00008967.1 26 IPI00307162.2

NH2-DNCCILDER-COOH Ace-ASLSLAPVNIFKAGADEER-COOH Ace-AAVKTLNPKAEVAR-COOH NH2-SASPM〈Mox〉 GVQDFDIVR-COOH NH2-PAKAAVHLEGKIEQAQR-COOH

16 IPI00220644.5 17 IPI00234667.2 18 IPI00383035.1

identified protein

40 KDA peptidyl-prolyl cis-trans isomerase ATP synthase R-chain, mitochondrial precursor capping protein (actin filament) muscle Z-line, R-2 C-myc binding protein cytokine receptor-like factor 3 fibrinogen β-chain precursor glutathione peroxidase 1 glyceraldehyde-3-phosphate dehydrogenase integrin-linked protein kinase 1 leukocyte elastase inhibitor platelet glycoprotein v precursor protein disulfide isomerase A3 precursor protein disulfide isomerase A5 precursor protein disulfide isomerase precursor putative ATP-dependent CLP protease proteolytic subunit, mitochondrial precursor pyruvate kinase 3 isoform 2 similar to bridging integrator 2 similar to fibrinogen, A R-polypeptide splice isoform 1 of P07226 tropomyosin R-4 chain splice isoform 1 of Q86UX7 UNC-112 related protein 2 splice isoform R-E of P02671 fibrinogen R/R-E chain, precursor splice isoform γ-B of P02679 fibrinogen γ-chain precursor T-complex protein 1, β subunit T-complex protein 1, ζ subunit thromboxane synthase VCL isoform meta-VCL

a The IPI database accession number of the identified proteins is indicated as well the measured N-terminal pyroglutamic acid, M〈Mox〉 ) methionine sulfoxide).

Concluding Remarks The 18O-peptide labeling method presented here differs from previously described procedures in two aspects. First, labeling takes place at moderately low pH-values. The rate of exchange is slower than at high pH but, importantly, hydrolytic activities are completely stopped. This principally avoids tryptic cleavage during the 18O-incorporation time, which may lead to incomplete labeling at new sites, complicating peptide identification and leading to incorrect quantitative assignments. The second important aspect of the procedure is the complete inactivation of trypsin by reductive alkylation under denaturing conditions. Although it is general practice to block trypsin hydrolytic activity by lowering the pH to 2, this is clearly not sufficient for the oxygen-exchange reaction. We found it necessary to destroy the trypsin structure by covalent modification of its thiol groups in order to avoid any re-naturation at every step of the subsequent peptide isolation procedure. This point was poorly addressed in previous studies and could therefore be accounted for as the major cause of reported variations in 18Oincorporation. Trypsin could also be blocked efficiently by performic acid oxidation27 but these conditions were found too acid, risking the initiation of the acid-catalyzed back exchange reaction. The procedure using 3 M guanidinium hydrochloride was therefore selected as the standard protocol. The labeling procedure is extremely simple and does not necessitate chemical reactions and subsequent removal of excess reagents by procedures that could result in important 790

Journal of Proteome Research • Vol. 3, No. 4, 2004

ratio

1.09 1.27 1.01 1.23 1.23 1.24 1.37 1.46 0.92 1.37 1.51 1.11 1.04 1.32 1.26 1.06 1.14 1.18 0.82 1.08 1.13 1.32 1.03 1.37 1.59 0.97 1.51 1.03 1.17 1.36 0.98

O:18O ratio (Ace ) acetylated N-terminus, Q〈Pyr〉 )

16

losses. In addition, the 18O-tagging step can be inserted at any stage in the course of different protocols in which differential analysis is anticipated. This is illustrated here, by the use of 18 O-labeling in a diagonal chromatography setup leading to the sorting of the protein NH2-terminal peptides of total cell lysates; a protocol in which at least two chromatographic steps and a chemical reaction are needed following the 18O-labeling step. It is important to state that no marked difference between the column retention characteristics of light peptides (16Olabeled) versus their heavy variants (18O-labeled) was noted. It has been reported in the past that polarity differences between C-H and C-D bonds cause peptides labeled with deuterated groups to elute earlier from reversed-phase columns compared to peptides that contain hydrogen. This hinders exact ratio determination since different ratios for one peptide couple are measured as a function of the time slice of the eluting peptides that is taken for determining the ratio.28 Recently, it was demonstrated that this isotope effect is not observed when 13Clabels are used.29 Since we also have not experienced this polarity/column retention effect using oxygen-18 isotopes we do believe that our strategy can be efficiently used for determining the exact ratios of eluting peptide couples. The 18O-tagging method which was optimized here for postcleavage labeling is not restricted to the use of trypsin. Other proteases such as the endoproteinases Lys-C and Glu-C were reported to equally incorporate two 18O-atoms during peptide hydrolysis.20 This is important, because it considerably broadens the application field, particularly in studies where multiple enzymes have to be used such as in post-translational modi-

research articles

Differential Oxygen-18 Incorporation in Proteomics

fication analysis in order to reach maximal sequence coverage, while conserving the differential labeling capacities of the procedure. In summary, post-cleavage trypsin-catalyzed oxygen-exchange could take a central position in nongel differential proteomics because it allows quantitative tagging of trypsincleaved peptides with 4 Da. This is a minimal mass difference in order to distinguish the 16O from the 18O isotope envelopes. We have demonstrated its general applicability in a study involving the isolation of the proteins NH2-termini as peptide acting as signatures for their parent proteins. Complex peptide mixtures with 16O- and 18O-labeled peptides were mixed and the ratios of the peptide, representing about 10% of the total amount, were calculated. This resulted in a symmetrical histogram from which the expected deviations can be easily calculated.

Acknowledgment. The authors thank Petra Van Damme, Sara De Groot, Koen Hugelier and Dr. Gre´goire R. Thomas for their fruitful discussions. K.G. is a Postdoctoral Fellow and L.M a Research Assistant of the Fund for Scientific ResearchFlanders (Belgium) (F.W.O.-Vlaanderen). The project was supported by research grants from the Fund for Scientific Research-Flanders (Belgium) (Project No. G.0008.03), the Inter University Attraction Poles (IUAP, Project No. P5/05) and the GBOU-research initiative (Project No. 20204) of the Flanders Institute of Science and Technology (IWT). References (1) Goshe, M. B.; Smith, R. D. Curr. Opin. Biotechnol. 2003, 14, 101119. (2) Lill, J. Mass Spectrom. Rev. 2003, 22, 182-194. (3) Kislinger, T.; Emili, A. Curr. Opin. Mol. Ther. 2003, 5, 285-293. (4) Oda, Y.; Huang, K.; Cross, F. R.; Cowburn, D.; Chait, B. T. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 6591-6596. (5) Conrads, T. P.; Alving, K.; Veenstra, T. D.; Belov, M. E.; Anderson, G. A.; Anderson, D. J.; Lipton, M. S.; Pasa-Tolic, L.; Udseth, H. R.; Chrisler, W. B.; Thrall, B. D.; Smith R. D. Anal. Chem. 2001, 73, 2132-2149. (6) Krijgsveld, J.; Ketting, R. F.; Mahmoudi, T.; Johansen, J.; ArtalSanz, M.; Verrijzer, C. P.; Plasterk, R. H.; Heck, A. J. Nat. Biotechnol. 2003, 21, 927-931.

(7) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Nat. Biotechnol. 1999, 17, 994-999. (8) Geng, M.; Ji, J.; Regnier, F. E. J. Chromatogr. A 2000, 870, 295313. (9) Munchbach, M.; Quadroni, M.; Miotto, G.; James, P. Anal. Chem. 2000, 72, 4047-4057. (10) Desiderio, D. M.; Kai, M. Biomed. Mass Spectrom. 1983, 10, 471479. (11) Mirgorodskaya, O. A.; Kozmin, Y. P.; Titov, M. I.; Korner, R.; Sonksen, C. P.; Roepstorff, P. Rapid Commun. Mass Spectrom. 2000, 14, 1226-1232. (12) Stewart, I. I.; Thomson, T.; Figeys, D. Rapid Commun. Mass Spectrom. 2001, 15, 2456-2465. (13) Yao, X.; Freas, A.; Ramirez, J.; Demirev, P. A.; Fenselau, C. Anal. Chem. 2001, 73, 2836-2842. (14) Liu, P.; Regnier, F. E. J. Proteome Res. 2002, 1, 443-450. (15) Bonenfant, D.; Schmelzle, T.; Jacinto, E.; Crespo, J. L.; Mini, T.; Hall, M. N.; Jenoe, P. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 880885. (16) Heller, M.; Mattou, H.; Menzel, C.; Yao, X. J. Am. Soc. Mass Spectrom. 2003, 14, 704-718. (17) Yao, X.; Afonso, C.; Fenselau, C. J. Proteome Res. 2003, 2, 147152. (18) Gaskell, S. J.; Haroldsen, P. E.; Reilly, M. H. Biomed. Environ. Mass Spectrom. 1988, 16, 31-33. (19) Takao, T.; Hori, H.; Okamoto, K.; Harada, A.; Kamachi, M.; Shimonishi, Y. Rapid Commun. Mass Spectrom. 1991, 5, 312315. (20) Schnolzer, M.; Jedrzejewski, P.; Lehmann, W. D. Electrophoresis 1996, 17, 945-953. (21) Gevaert, K.; Demol, H.; Verschelde, J. L.; Van Damme, J.; De Boeck, S.; Vandekerckhove, J. J. Protein Chem. 1997, 16, 335342. (22) Gevaert, K.; Goethals, M.; Martens, L.; Van Damme, J.; Staes, A.; Thomas, G. R.; Vandekerckhove, J. Nat. Biotechnol. 2003, 21, 566569. (23) Ferguson, P. L.; Smith, R. D. Annu. Rev. Biophys. Biomol. Struct. 2003, 32, 399-424. (24) Liener, I. E. J. Biol. Chem. 1957, 225, 1061-1069. (25) Hsu, J. L.; Huang, S. Y.; Chow, N. H.; Chen, S. H. Anal. Chem. 2003, 75, 6843-6852. (26) Michalet, S.; Favreau, P.; Stocklin, R. Clin. Chem. Lab. Med. 2003, 41, 1589-1598. (27) Hirs, C. H. W. J. Biol. Chem. 1956, 219, 611-621. (28) Zhang, R.; Sioma, C. S.; Wang, S.; Regnier, F. E. Anal. Chem. 2001, 73, 5142-5149. (29) Hansen, K. C.; Schmitt-Ulms, G.; Chalkley, R. J.; Hirsch, J.; Baldwin, M. A.; Burlingame, A. L. Mol. Cell. Proteomics 2003, 2, 299-314.

PR049956P

Journal of Proteome Research • Vol. 3, No. 4, 2004 791