Anal. Chem. 2006, 78, 6614-6621
Method for Quantitative Proteomics Research by Using Metal Element Chelated Tags Coupled with Mass Spectrometry Huiling Liu,†,‡,§,| Yangjun Zhang,†,‡,| Jinglan Wang,†,‡ Dong Wang,§ Chunxi Zhou,†,‡ Yun Cai,†,‡ and Xiaohong Qian*,†,‡
Department of Genomics and Proteomics, Beijing Institute of Radiation Medicine, 27 Taiping Road, Beijing 100850, China, Beijing Proteome Research Center, 33 Zhongguancun Life Park Road, Beijing 102206, China, and Department of Drug Metabolism and Pharmacokinetics, Beijing Institute of Transfusion Medicine, 27 Taiping Road, Beijing 100850, China
Proteomics, the analysis of the complete proteins of a cell or an organism, has grown rapidly as a subdiscipline of the life sciences.1 An important goal in proteomics is to compare differences in protein expressions in biological samples and attempt to correlate these differences with changes in physiological states
(quantitative or comparative proteomics).2 Currently, the two primary techniques used for relative quantitation in proteomics include two-dimensional gel electrophoresis (2-DE)-based methods and mass spectrometry (MS)-based methods with a stable isotope as the internal standard.1,2 2-DE, followed by MS analysis, has been used to separate complex protein mixtures and quantify protein spots.However, this method has limited dynamic range and often fails to detect lowabundance or membrane proteins effectively. 3 A modification of 2-DE, fluorescence 2-D difference gel electrophoresis (DIGE), uses different fluorescent dyes (Cy2, Cy3, Cy5) to label protein samples prior to 2-D PAGE. The DIGE technique allows multiple samples to be coseparated and visualized on a single 2-D gel.4 Although the dyes have a linear response to variation in protein concentration over 5 orders of magnitude, offer subnanogram sensitivity, and are compatible with MS analysis, the technique is compromised by high costs of both the necessary equipment and the expendable supplies such as the fluorescent dyes. Another commonly used technique for quantitative proteomics is based on stable isotope labeling followed by MS for quantitation and MS/MS for identification of a protein.5 Proteins under different physiological conditions are labeled with light and heavy reagents, respectively. After modification, both samples are mixed, reduced, alkylated, digested, and analyzed by MS and MS/MS. The light and heavy labeled peptides coelute in separation steps due to the similar chemical properties of isotopes. Quantitative analysis is performed by comparing the relative signal intensities of the light and heavy labeled peptides in the MS spectra; identification of the corresponding protein is obtained by selecting labeled peptides with higher signal intensity, then acquiring MS/ MS spectra, and finally searching them against protein databases. 6,7 Stable isotope labeling of the peptides or proteins can be accomplished by several methods. Stable isotopes can be intro-
* Corresponding author. E-mail:
[email protected]. Tel: +86-10-80705055. Fax: +86-10-8070-5155. † Beijing Institute of Radiation Medicine. ‡ Beijing Proteome Research Center. § Beijing Institute of Transfusion Medicine. | These authors contributed equally. (1) Righetti, P. G.; Campostrini, N.; Pascali, J.; Hamdan, M.; Astner, H. Eur. J. Mass Spectrom. 2004, 10, 335-348.
(2) Shao-en, O.; Matthias, M. Nat. Chem. Biol. 2005, 1, 252-262. (3) Aebersold, R.; Rist, B.; Gygi, S. P. Ann. N. Y. Acad. Sci. 2000, 919, 33-47. (4) Patton, W. F. J. Chromatogr., B 2002, 771, 3-31. (5) Leitner, A.; Lindner, W. J. Chromatogr., B 2004, 813, 1-26. (6) Washburn, M. P.; Wolters, D.; Yates, J. R. Nat. Biotechnol. 2001, 19, 242247. (7) Yates, J. R., III; Eng, J. K.; McCormack, A. L.; Schieltz, D. Anal. Chem. 1995, 67, 1426-1436.
The mass spectrometry-based methods with a stable isotope as the internal standard in quantitative proteomics have been developed quickly in recent years. But the use of some stable isotope reagents is limited by the relative high price and synthetic difficulties. We have developed a new method for quantitative proteomics research by using metal element chelated tags (MECT) coupled with mass spectrometry. The bicyclic anhydride diethylenetriamine-N,N,N′,N′′,N′′-pentaacetic acid (DTPA) is covalently coupled to primary amines of peptides, and the ligand is then chelated to the rare earth metals Y and Tb. The tagged peptides are mixed and analyzed by LC-ESI-MS/ MS. Peptides are quantified by measuring the relative signal intensities for the Y and Tb tag pairs in MS, which permits the quantitation of the original proteins generating the corresponding peptides. The protein is then identified by the corresponding peptide sequence from its MS/MS spectrum. The MECT method was evaluated by using standard proteins as model sample. The experimental results showed that metal chelate-tagged peptides chromatographically coeluted successfully during the reversedphase LC analysis. The relative quantitation results were accurate for proteins using MECT. DTPA modification of the N-terminal of peptides promoted cleaner fragmentation (only y-series ions) in mass spectrometry and improved the confidence level of protein identification. The MECT strategy provides a simple, rapid, and economical alternative to current mass tagging technologies available.
6614 Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
10.1021/ac060895j CCC: $33.50
© 2006 American Chemical Society Published on Web 08/17/2006
duced through in vivo metabolic labeling, such as SILAC, in which proteins are labeled by amino acids in cell culture,8,9 but in vivo metabolic labeling is sometimes not possible, as in human subjects. The stable isotopes can also be incorporated by using H218O during protein digestion.10-16 However, the trypsin-catalyzed and pH-mediated back exchange of 16O and 18O is a disadvantage of this approach. One or both carboxyl oxygens may be exchanged in this method, causing variability in the quantification. Last, stable isotope labeling can be realized by introducing a chemically stable isotope tag into a particular amino acid residue or the N-/Cterminus of proteins.17-23 The most popular example with this method is the isotope-coded affinity tag (ICAT).17 The ICAT reagent consists of a cysteine-specific reactive group. However, for some proteins that contain no or only one cysteine, ICAT is hardly useful or only used to quantify proteins on the basis of a single peptide, which will produce a result with great deviation. Due to the high price of stable isotope reagents used as the internal standards in the above-mentioned methods by mass spectrometric quantitation analysis, other proteomics quantitation methods were developed. Meares and colleagues designed an alternative method of protein labeling called element-coded affinity tag (ECAT)24based on tags containing different element-coded metal chelates that incorporate rare earth metals to the macrocycle 1,4,7,10-tetraazacyclododecane-N,N′,N′′,N′′′-tetraacetic acid (DOTA). The DOTA reagent specifically reacts with cysteine. In contrast to the ICAT method, which depends on biotin-avidin binding, the new ECAT method is based on interactions between an antibody and a range of ligands. When Meares et al. modified a synthetic peptide with either a carboxymethyl group or chelates of terbium or yttrium, the results suggested that the ECAT method had the potential to be used for relative protein quantitation in complex protein mixtures. However, the method has only been applied to standard peptides and not to proteins or proteomes. In contrast to modification of cysteine sulfhydryl groups in ECAT methods, we have developed a new method based on metal (8) Krijgsveld, J. Nat. Biotechnol. 2003, 21, 927-931. (9) Wu, C. C.; Maccoss, M. J.; Howell, K. E.; Matthews, D. E.; Yates, J. R., III. Anal. Chem. 2004, 76, 4951-4959. (10) Mirgorodskaya, O. A.; Kozmin, Y. P.; Titov, M. I.; Korner, R.;.Sonksen, C. P.; Roepstorff, P. Rapid Commun. Mass Spectrom. 2000, 14, 1226-1232. (11) Yao, X.; Freas, A.; Ramirez, J.; Demirev, P. A.; Fenselau, C. Anal. Chem. 2001, 73, 2836-2842. (12) Staes, A.; Demol, H.; Van Damme, J.; Martens, L.; Vandekerckhove, J.; Gevaert, K. J. Proteome Res. 2004, 3, 786-791. (13) Yao, X.; Afonso, C.; Fenselau, C. J. Proteome Res. 2003, 2, 147-152. (14) Bantscheff, M.; Dumpelfeld, B.; Kuster, B. Rapid Commun. Mass Spectrom. 2004, 18, 869-876. (15) Brown, K, J.; Fenselau, C. J. Proteome Res. 2004, 3, 455-462. (16) Hicks, W. A.; Halligan, B. D.; Slyper, R. Y.; Twigger, S. N.; Greene, A. S.; Olivier, M.J. Am. Soc. Mass Spectrom. 2005, 16, 916-925. (17) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Nat. Biotechnol. 1999, 17 (10), 994-999. (18) Ji, C.; Guo, N.; Li, L. J. Proteome Res. 2005, 4 (6), 2099-2108. (19) Zhang, R.; Sioma, C. S.; Wang, S.; Regnier, F. E. Anal. Chem. 2001, 73 (21), 5142-5149. (20) Zhou, H.; Ranish, J. A.; Watts, J. D.; Aebersold, R. Nat. Biotechnol. 2002, 20 (5), 512-515. (21) Li, J.; Steen, H.; Gygi, S. P. Mol. Cell. Proteomics 2003, 2 (11), 1198-1204. (22) Lu, Y.; Bottari, P.; Turecek, F.; Aebersold, R.; Gelb, M. H. Anal. Chem. 2004, 76 (14), 4104-4111. (23) Ross, P. L.; Huang, Y. N.; Marchese, J. N.; Williamson, B.; Parker, K.; Hattan, S.; Khainovski, N.; Pillai, S.; Dey, S.; Daniels, S.; Purkayastha, S.; Juhasz, P.; Martin, S.; Bartlet-Jones, M.; He, F.; Jacobson, A.; Pappin, D. J. Mol. Cell. Proteomics 2004, 3, 1154-1169. (24) Whetstone, P. A.; Butlin, N. G.; Corneillie, T. M.; Meares, C. F. Bioconjugate Chem. 2004, 15, 3-6.
chelates and labeling of primary amines in peptides for relative quantification and identification, which is called metal element chelated tags (MECT). To introduce the rare earth metal chelate tag, the bicyclic anhydride diethylenetriamine-N,N,N′,N′′,N′′pentaacetic acid (DTPA) is covalently coupled to the peptide amine and then chelated with the rare metal. After the peptides are labeled with different rare metal ions of Y and Tb, the tagged peptides are mixed and analyzed by LC-MS/MS. Proteins are quantified by ion density ratio in the MS spectra and identified by MS/MS spectra followed by protein database searching. The strategies developed here have several advantages. First, DTPA dianhydride is an inexpensive and easily obtained reagent and has been commonly used for the covalent modification of proteins. Second, as opposed to the cysteine sulfhydryl group modification in ECAT methods, the reagent labels all primary amines regardless of peptide classes, which theoretically makes the approach capable of quantifying all the proteins generating every observed peptide. Third, a DTPA-modified N-terminal of peptides could promote cleaner fragmentation (only y-series ions) and increase the confidence level of protein identification. Finally, because the eight rare earth elements are naturally monoisotopic and can produce 28 different mass tag pairs depending on the varieties of permutation and combination, the selective scope of possible internal standards can be enlarged in proteomics quantitation. EXPERIMENTAL SECTION Chemicals and Materials. Diethylenetriamine-N,N,N′,N′′,N′′pentaacetic acid dianhydride (DTPAA) was purchased from Dojindo (Kumamoto, Japan). YCl3, TbCl3, triethylammonium bicarbonate (TEAB), and tris(2-carboxyethyl)phosphine (TCEP) were obtained from Sigma (St. Louis, MO). All other chemicals used were of analytical grade or the highest quality that was commercially available. HPLC grade acetonitrile was acquired from J. T. Baker. Deionized water was produced by a Milli-Q A10 system from Millipore (Bedford, MA). Preparation and Digestion of Standard Protein Mixture. Bovine serum albumin (BSA), R-lactalbumin from bovine milk, β-lactoglobin from bovine milk, myoglobin (nonapo form) from equine skeletal muscle, lysozyme from chicken egg white, bovine apotransferrin, and bovine insulin were purchased from SigmaAldrich (St. Louis, MO) and used without further purification. Two protein mixtures were prepared in-house. The first sample was a mixture containing equimolar amounts of BSA, R-lactalbumin, β-lactoglobin, myoglobin, lysozyme, and apotransferrin (1 nmol of each protein). The second sample included two mixtures containing the same six proteins but at different concentrations (µg/mL in mixtures A and B, BSA (2, 10), R-lactalbumin (5, 2), β-lactoglobin (10, 10), myoglobin (4, 2), lysozyme (2, 5), and apotransferrin (10, 2)). Protein mixtures were buffered at pH 7.0 with 100 mM TEAB, denatured by heat up to 95 °C for 5 min, reduced with 2 µL of 50 mM TCEP at 60 °C for 1 h, alkylated with 1 µL of 500 mM iodoacetamide at 37 °C for 1 h in the dark, and digested with trypsin (1:20 w/w) at 37 °C for 18 h. Peptide Coupling with DTPAA and Then Chelation of Rare Earth Metals. The solution of the trypsin-digested peptides (20 nmol/µL) buffered at pH 7.0 with 50 mM bicarbonate was added to the solid anhydride (typically 360 µg), vortexed to mix for 1 min, then centrifuged to collect the sample at the bottom of the Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
6615
tube, and reacted for 30 min at room temperature. Peptides may be coupled with DTPA simply by reaction with the cyclic anhydride in aqueous solution with high efficiency.25 After coupling, the sample was dried in a centrifugal vacuum concentrator, reconstituted with equal volume of 0.1 M NH4OAc buffer at pH 6.0, combined with 2-fold molar excess amounts of either TbCl3 or YCl3 solutions, and then incubated at 37 °C for 2 h. After the addition of YCl3 or TbCl3 to the solution of peptide-DTPA compounds, the labeling process was quantitatively completed under the experimental conditions mentioned above. The solutions of Y and Tb labeled peptides were combined in various proportions and analyzed by LC-MS/MS as described below. MALDI-TOF Mass Spectrometry Analysis. MALDI-TOF mass spectra were acquired on a 4700 Proteomics Analyzer (Applied Biosystems, Framing, MA). Peptide samples were directly mixed at 1:9 (v/v) with saturated R-cyano-4-hydroxycinnamic acid (5 mg/mL) in 50% acetonitrile containing 0.1% trifluoroacetic acid, and 0.5 µL was spotted onto target plates. Samples were air-dried, and MS spectra were obtained in positive reflection mode at an accelerating voltage of 20 kV. All m/z values quoted in this study are monoisotopic. Quantitation of Proteins by LC-ESI-Q-TOF. The eluted peptide mixtures were immediately analyzed by an ESI-Q-TOF mass spectrometer after separation on the coupled CapLC system (Waters). The mass spectrometer was operated in the positive survey mode and externally calibrated with fragments of Glufibrinopeptide. The instrument conditions were as follows: electrospray voltage at 3.2 kV, sample cone voltage at 45 V, and source temperature at 80 °C. The flow from pump C was used to load the sample and wash the sample onto a 320 µm i.d. × 1 mm precolumn (LC Packings) for 4 min with a 0.1% formic acid aqueous solution. Separation was performed on a 75 µm i.d. × 15 cm PepMap C18 column (LC Packings) using a solvent gradient system (solution A, 0.1% formic acid in water/acetonitrile, 95/5, v/v; solution B, 0.1% formic acid in water/acetonitrile 15/85, v/v). The flow from the pumps was 3.0 µL/min and was split to 0.3 µL/min by a glass capillary. Gradient elution was as follows: 4% buffer B, over 4.5 min; 4% to 50% buffer B over 45 min; 50 to 100% buffer B over 10 min; 100% buffer B over 10 min. The ratios of the Y and Tb element-labeled peptides were calculated from their ion intensity. The relative quantitation of a protein from two different samples was determined by averaging the ratios of the Y and Tb element-labeled peptides that were unique to the same protein. Database Searching. Peptide and protein were identified by submitting product ion spectra of the peptides to the Uniprot-SP (version 2.0); Mascot search engine (version 1.9) was used to search the protein database.26 Searched parameters included a modification of +465 Da to DTPA-Y and +535 Da to DTPA-Tb (mass addition due to modification by DTPA coupling metals element) at primary amines (N-terminal and lysine). Modifications to the above parameters included selection of methionine oxidation as variable modifications and alkylation by iodoacetamide as fixed modifications. One missed cleavage was allowed and precursor mass tolerance less than 50 ppm. (25) Meares, C. F.; Wensel, T. G. Acc. Chem. Res. 1984, 17, 202-209. (26) Perkins, D. N.; Pappin, D. J.; Creasy, D. M.; Cottrell, J. S. Electrophoresis. 1999, 20 (18), 3551-3567.
6616 Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
Figure 1. Schematic diagram of DTPA coupled with peptide and chelated with metal ions. “P” represents the peptide and “M” represents the rare metal ion Y or Tb. The mass difference between peptides with Y and Tb tags is 70 Da.
RESULTS AND DISCUSSION Principles of the Metal Element Chelated Tags. Apart from stable isotopes as internal standards, rare metal elements, chemically similar with each other, are also proper internal standards for quantitative proteomics based on mass spectrometry. Rare earth elements could be used as mass tags in different combinations from 1 to 86 Da. 24 Most of the rare earth elements are naturally monoisotopic, thus providing a variety of simple choices for preparing mass tags. The schematic diagram of DTPA coupled to peptide and a chelated metal ion is shown in Figure 1. According to the MECT method, the proteins were first digested by trypsin; then DTPAA specifically coupled to primary amines of peptides; afterward, the peptides chelated with rare metal ions in respective buffer. The chosen metal ions in this experiment were Y and Tb. The peptideDTPA chelates binding metal ions Y and Tb with one charge differing by 70 Da, large enough to avoid the overlapping of mass to charge peaks for the Tb and Y tag labeled peptides. Besides, the mass difference between Y and Tb tagged peptides cannot be mistaken for other common mass differences such as m/z at 6 for water loss with charge 3+, 7 for methylene group with charge 2+, 8 for methionine oxidation with charge 2+, and 9 for water loss with charge 2+. Reactivity of Primary Amines of Peptides in MECT Method. To test the reactivity of DTPAA with the primary amines of peptides, a model protein, myoglobin, was selected and digested by trypsin first to produce a mixture of peptides. The digest was then reacted with DTPAA in TEAB buffer. The resulting sample was analyzed by MALDI-TOF MS as shown in Figure 2. Generally, the amino terminus and internal lysine of each peptide could be labeled, thus enhancing peptide coverage for any given protein to be identified. The modification of DTPA added 375.1 Da to the most parent ions of the myoglobin peptides, such as peptides with m/z 748.44 and 1606.88 increased to m/z 1123.59 and 1982.1, respectively; peptides with m/z 1859.85 and 2174.11 were the ions after neutral loss of m/z 1502.69 and 1815.93 modified by DTPA. Although most of the peptides coupled with DTPA at an N-terminal of the peptide, a few peptides coupled with DTPA at the Lys side chain at the same time, such as peaks m/z 1271.68 and 1378.85 producing peaks at m/z 2021.88 and 2029.05 in Figure 2. All of the primary amines of peptides would be modified by the DTPA, and this is a common practice in using amine-modifying tags in proteomics. The standard peptide LRRASLG (kemptide acetate salt; Sigma) was coupled with DTPA and then chelated with Y and Tb ion in Figure 3. Kemptide labeled with either Y or Tb tag was finally mixed in a ratio of 1:1. The top panel A depicts the peak of
Figure 2. MALDI-TOF MS spectrum of the reaction product of the DTPA coupling with myoglobin peptides. The spectrum of the reaction product was obtained after 1-h reaction. The top panel depicts the m/z values of myoglobin peptides digested by trypsin. The bottom panel shows the m/z values of the myoglobin peptides covalently coupled with DTPA (marked by the arrows).
Figure 3. MALDI-TOF MS spectrum of standard peptide kemptide (772 m/z) coupled and labeled by DTPA and Y/Tb. The top panel A is the MALDI-TOF MS spectrum of the reaction product of the DTPA coupling with standard peptide kemptide (772 m/z). The bottom panel B shows the m/z value of the kemptide peptides coupling with DTPA and chelating with Y/Tb ions mixed in the ratio of 1:1.
kemptide (772.4 Da) coupled with DTPA (375.1 Da), whose m/z was 1147.5. The bottom panel shows the m/z values of the kemptide peptides coupling with DTPA (1147.5 Da) and chelating with Y/Tb ion mixed in the ratio of 1:1. The value of Y tag and Tb tag of kemptide was 1233.2 and 1303.2 Da, respectively, which differs by 70 Da as expected. The reaction of DTPAA with amino group of peptides reached completion between 0.5 and 1 h, and the product remained stable for at least one week. We chose 1 h as the incubation time for all
subsequent experiments. A 50-fold molar excess of DTPAA with respect to the protein was used. From the mass spectrum in Figure 3A, the parent peptide of 772.4 Da was completely converted to 1147.5 Da, so the reaction was specific and highly efficient. This result was consistent with that of Hnatowich and colleagues.27 Then half of the product was reacted with YCl3 and the other half with TbCl3. The labeling by chelating with Y or Tb (27) Hnatowich, D. J.; Layne, W. W.; Childs, R. L.; Lanteigne, D.; Davis, M. A. Science 1983, 220, 613-615.
Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
6617
Figure 4. Spectra of Y-DTPA-LRRASLG and Tb-DTPA-LRRASLG at different times (2 h, 12 h, left panel) and in different solutions (mobile phases A and B, right panel). Panels A and C are the spectra of equimolar TbCl3 solution added into the Y-DTPA-LRRASLG solution and panels B and D are the spectra of equimolar YCl3 added into the Tb-DTPA-LRRASLG solution. Panels E-H are the spectra of the two tagged peptides incubated in mobile phase A and mobile phase B for 12 h at room temperature.
was extremely rapid and was, in effect, completed upon mixing YCl3 or TbCl3 with solutions of DTPA-peptides (see Figure 3B). Investigation on the Stability of Metal-DTPA-Peptide. In principle, metal-DTPA-peptide complexes exist in equilibrium with free metal ions in solution, potentially causing scrambling between metal ions and partially contributing to biased observed ratios. To address this issue, we have investigated the stability of the metal-DTPA-peptide complex formed with the standard peptide LRRASLG. (Figure 4). First, we added equimolar amounts of TbCl3 solution into the Y-DTPA-LRRASLG solution and equimolar amounts of YCl3 solution into the Tb-DTPA-LRRASLG solution, respectively. Then the two solutions were placed at room temperature for 12 h (Figure 4A-D) or placed in the mobile phase A and mobile phase B at room temperature for 12 h (Figure 4E, F). Despite a little interference (marked with an ellipse), the spectra of Y-DTPA-LRRASLG and Tb-DTPALRRASLG indicate that both labeled peptides were relatively stable in solutions of either the other rare earth metal or in mobile phase. Chromatographic Performance and CID Behavior of Tagged Peptides. Figure 5A shows the total ion current chromatograms of two peptides from insulin (GFFYTPK and FVNQHLCGSHLVEALYLVCGER) mixed in the same ratio of 1:2 with different tags. As can be seen in Figure 5A, tagged peptides containing different rare earth elements coeluted on reversedphase liquid chromatography at 29.87 (GFFYTPK) and 31.32 min (FVNQHLCGSHLVEALYLVCGER), respectively. A full scan of the mass spectrometer operated in MS mode is shown in panel B of Figure 5. The pair of Y and Tb labeled peptides (GFFYTPK) was detected and coeluted from the LC column at 29.87 min with respective m/z values, which differed by m/z 35 for the ion pair with a charge of 2. The mass difference between these two ions in the pair of peptides depended on the charge state (number of hydrogen ions) and was typically either m/z 35 or 70. 6618 Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
Panel C in Figure 5 shows the intensity ratio of observed Y /Tb labeled peptide GFFYTPK with charge +2. The m/z values for Y tag and the Tb tag were 660.58 and 695.55, respectively. The LC-MS analysis was repeated 6 times, and the average of observed Y /Tb labeled peptide ratio was 0.38 ( 0.02 (n ) 6). The intensity ratio observed was unequal to the expected ratio (0.5), which may be caused by an ionization suppression effect and will be discussed later. From Figure 6, the tandem MS spectrum of the precursor ion, Tb labeled peptide (M + 3H)3+, FVNQHLCGSHLVEALYLVCGER, produced predominantly y-series ions in the MS/MS spectrum, and the tags themselves did not complicate mass spectrometric analysis. Database searching with this tandem spectrum showed that the protein is insulin. Similar results were also obtained for peptides with Y tags. The DTPA-derived peptides facilitated cleaner fragmentation, mainly resulting in y-series ions in the MS/ MS spectra. The confidence level of peptide identification can be enhanced by a single y-type ion originating from one parent peptide according to Lee et al.28,29 Verification of the Ionization Suppression Effect. To determine whether the ionization suppression effect is affected by different peptides, a protein mixture was prepared containing six proteins in equal molar concentrations. The mixture was reduced, alkylated, digested, and labeled with Y or Tb tags as described in the Experimental Section. Then, both Y and Tb labeled peptide mixtures of apotransferrin, BSA, myoglobin, β-lactoglobin, R-lactalbumin, and lysozyme were mixed in the ratio of 1:1. Aliquots (∼50 ng) of both mixtures were analyzed by LCMS/MS. The identification (amino acid sequences) of the peptides was performed by searching MS/MS spectra in a separate LC(28) Lee, Y. H.; Kim, M. S.; Choie, W. S.; Min, H. K.; Lee, S. W. Proteomics 2004, 4, 1684-1694. (29) Goodlett, D. R.; Bruce, J. E.; Anderson, G. A.; Rist, B.; Pasa-Tolic, L.; Fiehn, O.; Smith, R. D.; Aebersold, R. Anal. Chem. 2000, 72, 1112-1118.
Figure 5. Analysis of the labeled peptide mixture of insulin by LC-MS. Prepared molar ratio of peptides with Y tag to Tb tag was 1:2. The bottom of panel A depicts the relative intensity versus time; the top four panels of (A) are the reconstructed ion chromatograms for the Y and Tb tagged peptides, respectively, showing the elution times for tagged peptides containing different rare earth elements on reversed-phase liquid chromatography. Mass spectral analysis of the Y and Tb labeled peptides of insulin (GFFYTPK) in panels B and C. Panel B is the spectrum of the Y and Tb labeled peptide of insulin (GFFYTPK), which differs by 35 mass units. The peptides have been coeluted from the LC column at 29.87 min. The ion intensity ratio of observed Y/Tb tag is presented in panel C.
MS/MS analysis against the Swissprot Protein database using the Mascot database searching software. Figure 7A shows the distribution of the theoretical and observed ratios of Y /Tb labeled peptides in six protein mixtures. All observed Y /Tb tag ratios were approximately 0.65, although the value was not consistent with the theoretical ratio of 1:1 for the six proteins. This phenomenon indicates that the ionization suppression effect was identical for any peptide and, therefore, for the corresponding proteins. The average Y /Tb tag ratios for apotransferrin, BSA, myoglobin, β-lactoglobin, R-lactalbumin, and lysozyme peptides were 0.69 ( 0.02 (peptide number n ) 13), 0.68 ( 0.04 (n ) 11), 0.67 ( 0.07 (n ) 9), 0.695 ( 0.09 (n ) 7), 0.71 ( 0.13 (n ) 3), and 0.66 ( 0.11 (n ) 4), respectively. The average and the standard deviation values were calculated using different peptides belonging to the same protein. Dynamic Range of the MECT Method. To evaluate the feasibility of the MECT strategy for comparative proteomics, insulin was reduced by TCEP, digested by trypsin, reacted with DTPAA, and separately chelated with either YCl3 or TbCl3, as described in the Experimental Section. Then the different ratios
of Y and Tb labeled peptides, mixed in ratios 1:1, 2:1, 5:1, 10:1, 1:2, 1:5, and 1:10, were subjected to the separation of a 60-min nanoLC gradient run on CapLC followed by the analysis of ESI Q-TOF MS with data-dependent mode. Figure 7B demonstrates the linear change of observed and theoretical ratios of Y/Tb labeled peptide. The correlation coefficients (r2) of the linear regression lines for the two peptides were 0.99981 for GFFYTPK and 0.99953 for FVNQHLCGSHLVEALYLVCGER, respectively, indicating good linearity with respect to Y/Tb labeled peptide ratios over the range of 0.1-10. The RSD values of the Y/Tb labeled peptide ratio were less than 10%, suggesting the accuracy of the method. The observed linear response suggests that MECT methods could be useful for measuring ratios in more complex mixtures. Application of the MECT Method to Standard Protein Mixture. To further demonstrate the sensitivity of this approach for proteomics applications, another pair of mixtures A and B consisting of the same six proteins at known but different concentrations was prepared and analyzed with the same method as above. The preparation of mixtures A and B is referred to in Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
6619
Figure 6. Tandem MS spectrum of the precursors, Tb labeled peptide (M + 3H)3+ generated from a Q-TOF mass spectrometer (peptide sequence: FVNQHLCGSHLVEALYLVCGER, m/z 2601.2). Fragment ions in the y-series were labeled. Similar results were obtained for peptides with Y tags.
Figure 7. Distribution of the theoretical and observed ratios of Y/Tb labeled peptides in six protein mixtures (A, left). The peptide mixture of apotransferrin, BSA, myoglobin, β-lactoglobin, R-lactalbumin, and lysozyme in equal molar concentration was labeled by YCl3 or TbCl3, respectively, and mixed in 1:1 ratio, and then a 50-ng sample was analyzed by LC-MS. The average ratios of the Y/Tb tag calculated from the observed peak intensities of the peptides are plotted. The straight dashed line represents the theoretical Y/Tb tag ratios. Panel B (right) is plots of the observed ratios (Y/Tb labeled) of the two peptides of insulin versus their theoretical ratios. Linear regression lines for theoretical ratios of Y/Tb tags (9), peptides with Y/Tb tags: GFFYTPK (b) and FVNQHLCGSHLVEALYLVCGER (2) are obtained from six different ratios ranging from 0.1 to 10 for each peptide. The correlation coefficients (r2) of the linear regression lines for the two peptides are presented along with the corresponding peptide lines. The vertical bar at each point indicates the RSD of the six repeated LC-MS analyses.
the Experimental Section. Mixture A was tagged with the Y and mixture B with Tb. The tagged peptide mixture (10 µL, corresponding to 300 fmol of BSA in mixture A and 260 fmol of apotransferrin in mixture B) was quantified and sequenced by LC-ESI Q-TOF mass spectrometer. To offset the ion suppression effect, a calibration curve was prepared using myoglobin digested peptides and a linear equation (y ) 0.050665 + 0.65201x) was calculated and used to rectify the observed ratios of Y/Tb tags. For all primary amines of peptides 6620 Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
labeled by using the mass tagging method, the overall coverage of proteins was improved, and the confidence level of protein identification was increased by aligning the MS/MS spectra of serial y ions from the DTPA-modified peptides with the protein database. All six proteins were unambiguously identified and accurately quantified (Table 1) with multiple tagged peptides matched to each protein. The mean deviation between the observed and expected quantities for the six proteins ranged from 1.27 to 9.76%. The results show that this technique can accurately
Table 1. Metal Element Chelated Tags Peptide Identified from Protein Digest Mixture by QTOF MS
protein
peptide sequence
apotransferrin
DGAGDVAFVK STIFENLANK AIAANEADAV LDAGLVYDAY APNNLKPVVA FYGSK NLNEKDYELL LDGTR EDLIWELLNQ QEHFGK KPVEEYANCH AR KCSTSSLLEAC FR ADRDQYELLC DNTR DCHLAQVPSH VVAR LGEYGFQNAL VR KVPQVSTPTLV VSR FKDLGEEHFK CCAADDKEAC AVEGPK KQTALVELLK VEADIAGHGQ VLIR TEAEMK YKELGFQG GHHEAELKPL QSHATK VAGTWYSLA AASDISLLDAQ APLR FDKALKALPM IR TKIPAVFKIDA NENK EQLTKCEVFR VGINYWLAHK ILDKVGINYW AHK FESNFNTQATNR NTDGSTDYGI QINSR CKGTDVQAWIR
bovine serum albumin
myoglobin
β-lactoglobulin
R-lactalbumin
lysozyme
ratio (Y /Tb tag) theor rectifieda
peptide mass
obsd
2233.1
3.32
3954.0
3.33
5.03
1895.9
3.31
5.00
2070.0
3.30
4.98
1529.8
3.45
5.21
1545.7
3.34
5.04
1824.9
3.43
5.18
1632.8
3.39
5.12
1479.8
0.17
1639.9
0.15
0.15
1249.6 1756.7
0.17 0.18
0.18 0.20
1142.7 1606.9
0.18 1.38
708.3 941.5 1854.0
1.36 1.39 1.38
2707.4
0.68
1539.8
0.76
1.09
1801.0
0.73
1.04
1252.6 1200.6 1669.9
1.64 1.66 1.68
2.5
1428.7 1753.8
0.33 0.36
0.4
1276.6
0.32
measure the relative quantitation of proteins in mixtures and provides a simple, rapid, and economical alternative to currently available mass tagging technologies. CONCLUSION A new strategy was proposed for quantitative analysis of complex protein mixtures based on rare earth metal labeling, during which bicyclic anhydride DTPA was first covalently coupled to peptides and then chelated with a rare earth metal. Rare earth elements that bind to DTPA could extend the choice dramatically, and their mass defects give the masses of tagged peptides exact values not normally shared by molecules that contain only light elements. Other advantages of this approach include the convenient and rapid sample preparation and good sensitivity at the femtomole level, which could meet the requirements of proteomics research potentially for biological samples. Importantly, all primary amines were labeled, and the cleaner fragmentations increased the overall coverage and the confidence level of protein identification. It needs to be pointed out that, in the above experiment,
5.0
0.2
2.0
5.01
0.18
0.20 2.04
mean ( SD
% error
5.07 ( 0.09
1.40
0.18 ( 0.02
10.0
2.04 ( 0.02
2.00
1.03 ( 0.06
3.00
2.44 2.47 2.50
2.47 ( 0.03
1.20
0.43 0.47
0.44 ( 0.03
10.00
2.01 2.05 2.04 1.0
0.97
0.41
although the average observed ratio for standard proteins was 0.65 by the method, the deviation of relative quantitation caused by ionization suppression effect can be rectified by multiplying by the rectifying coeffient. While there are some limitations inherent in this method, such as elaborate metal chelating chemistry and high molecular weight of the modified products, it has promising application to complex biological samples for relative quantitative analysis in proteomes. ACKNOWLEDGMENT The authors received financial support for the work from National Key Program for Basic Research (2001CB510201, 2004CB518707), National Natural Science Foundation (30321003, 20405017, 20505018, 20505019), and Beijing Municipal Program for Science & Technology (H030230280190). Received for review May 16, 2006. Accepted July 21, 2006. AC060895J Analytical Chemistry, Vol. 78, No. 18, September 15, 2006
6621
