Fluorescence-Based Peptide Labeling and Fractionation Strategies for

Fluorescence-Based Peptide Labeling and Fractionation Strategies for Analysis of Cysteine-Containing Peptides. Adrienne Clements, Murray V. Johnston, ...
0 downloads 0 Views 214KB Size
Anal. Chem. 2005, 77, 4495-4502

Fluorescence-Based Peptide Labeling and Fractionation Strategies for Analysis of Cysteine-Containing Peptides Adrienne Clements,† Murray V. Johnston,† Barbara S. Larsen,‡ and Charles N. McEwen*,‡

Department of Chemistry, University of Delaware, Newark, Delaware 19716, and DuPont Experimental Station, DuPont Corporate Center for Analytical Sciences, Wilmington, Delaware 19880-0228

This study demonstrates that 1,5-I-AEDANS (5-({2[(iodoacetyl)amino]ethyl}amino)naphthalene-1-sulfonic acid) can be used as a versatile fluorescence-based peptide quantification tool and provides readily interpretable tandem mass spectra for de novo peptide sequencing. Two AEDANS-cysteinyl-peptide fractionation strategies were evaluated. One AEDANS-cysteinyl-peptide fractionation strategy employs immobilized metal affinity chromatography (IMAC) to recover AEDANS-labeled peptides and reduce the complexity of peptide mixtures. In an alternate solid-phase approach, 1,5-I-AEDANS was coupled to an o-nitrobenzyl-based photocleavable resin to produce a resin that can label and isolate thiols and cysteinecontaining peptides with a modified-AEDANS label (mAEDANS: 5-((4-amino-4-oxobutanoyl){2-[(iodoacetyl)amino]ethyl}amino)naphthalene-1-sulfonic acid). This fractionation protocol enriches cysteine-containing peptides more specifically than the IMAC strategy. Using microLC-ESI-MS with an on-line fluorescence detector and a Q-TOF mass spectrometer, we generated fluorescencebased elution profiles and corresponding positive ion mass spectra of AEDANS-labeled peptides. This study demonstrates that AEDANS-peptides produce positive ion ESI-MS mass spectra with detection limits comparable to those of the unlabeled peptide. Collision-induced dissociation (CID) of fluorescent AEDANS-peptides revealed readily interpretable product ion spectra with the label intact. Similar to the AEDANS-labeled peptide, an mAEDANS-labeled thiol is fluorescent and CID of a mAEDANS-labeled peptide also reveals an interpretable product ion spectrum with the label intact. With the maturation of the genomic era, a variety of tools have been developed to quantify protein levels from complex biological samples (for review, see ref 1). The primary goal of these tools is to determine differences in protein levels from cells (or tissues) with differing physiological states. A gel-based quantification strategy is the traditional approach in which two-dimensional * Corresponding author. Phone: (302) 695-2952, Fax: (302) 695-1351. E-mail: [email protected]. † University of Delaware. ‡ DuPont Corporate Center for Analytical Sciences. (1) Julka, S.; R. F. J. Proteome Res. 2004, 3, 350-363. 10.1021/ac050247k CCC: $30.25 Published on Web 06/21/2005

© 2005 American Chemical Society

polyacrylamide gel electrophoresis (2D-PAGE) is employed to separate proteins based on isoelectric points and molecular masses.2 Visualized spots on the gel correspond to proteins, and spot intensity is used to determine the relative protein expression levels from samples. Proteins of interest are then digested by proteases and extracted from the gel spots for identification using tandem mass spectrometry.3-5 Although this strategy is theoretically straightforward, it is often technically challenging to obtain reproducible spot profiles from one gel to the next. Consequently, difference gel electrophoresis was developed, which employs spectrally distinct Cy fluorescent dyes to label proteins from different samples.6-8 The differentially labeled proteins are mixed, and 2D-PAGE of the mixture reveals altered protein expression levels based on the comparison of fluorescent levels from these distinguishable protein populations, allowing direct comparison of the protein samples in one gel. Other strategies have been developed for peptide quantification using LC-MS techniques. These strategies often require less intensive sample preparation and generally employ isotopic labeling as a tool to compare protein levels of two complex biological samples using mass spectrometry.1 Isotope labels can be introduced into proteins in a variety of ways, ranging from metabolic incorporation8-11 to postbiosynthesis covalent modifications. The former is generally most practical for samples derived from cell cultures, while the latter can be applied to a wide variety of sample sources. Postbiosynthesis strategies include enzymatic 18O incorporation using endoproteases12-18 and chemical labeling (2) Klose, J.; K. U. Electrophoresis 1995, 16, 1034-1059. (3) Page, M. J.; A. B.; Townsend, R. R.; Parekh, R.; Herath, A.; Brusten L.; Zvelebil, M. J.; Stein, R. C.; Waterfield, M. D.; Davies, S. C.; O’Hare, M. J. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 12589-12594. (4) Srisomsap, C.; S. P.; Otto A.; Mueller, E. C.; Punyarit, P.; Wittmann-Liebold, B.; Svasti, J. Proteomics 2002, 2, 706-712. (5) Lahm, H. W.; L. H. Electrophoresis 2000, 21, 2105-14. (6) Tonge, R.; S. J.; Middleton, B.; Rowlinson, R.; Rayner, S.; Young, J.; Pognan, F.; Hawkins, E.; Currie, I.; Davison, M. Proteomics 2001, 1, 377-396. (7) Shaw, J.; R. R.; Nickson, J.; Stone, T.; Sweet, A.; Williams, K.; Tonge, R. Proteomics 2003, 3, 1181-1195. (8) Washburn, M. P.; U. R.; Deciu, C.; Schieltz, D. M.; Yates, J. R., 3rd. Anal. Chem. 2002, 74, 1650-1657. (9) Oda,Y.; H. K.; Cross, F. R.; Cowburn, D.; Chait, B. T. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 6591-6596. (10) Pasa-Tolic, L.; J. P.; Anderson, G. A.; Lipton, M. S.; Peden, K. K.; Martinovic, S.; Tolic, N.; Bruce, J. E.; Smith, R. D. J. Am. Chem. Soc. 1999, 121, 79497950. (11) Smith, R. D.; A. G.; Lipton, M. S.; Pasa-Tolic, L.; Shen, Y.; Conrads, T. P.; Veenstra, T. D.; Udseth, H. R. Proteomics 2002, 2, 513-523. (12) Reynolds, K. J.; Y. X.; Fenselau C. J. Proteome Res. 2002, 1, 27-33.

Analytical Chemistry, Vol. 77, No. 14, July 15, 2005 4495

(for review, see ref 1). The isotope coded affinity tag (ICAT) system is perhaps the most widely recognized peptide labeling strategy. The ICAT strategy covalently modifies cysteine residues of proteins from two samples with isotopically heavy or light tags containing a biotin moiety.19-21 The labeled protein samples are mixed and proteolyzed. The ICAT-labeled peptides can be isolated using avidin-based chromatography to simplify the peptide mixture. The relative quantities of isotopically heavy and light peptides correspond to relative protein abundances from the two samples and can be determined using LC-MS. Another approach to protein quantification utilizes tandem mass tags, which uses isobaric peptide labeling reagents that yield signature ions in the low-mass regions of collision-induced dissociation (CID) product ion spectra.22 The relative intensities of these signature product ions can be used to compare peptide quantities. In general, mass spectrometry is used in proteomic applications for protein/peptide quantification using MS (as mentioned above with standard isotope labeling strategies) and protein/peptide identification using tandem MS for de novo sequencing. It is difficult to get quantitative and sequencing information during a single experiment as a mass spectrometer switches acquisition modes from MS to MS/MS. Therefore, a sensitive spectroscopic technique, such as on-line fluorescence detection, may be a suitable alternative to MS-based quantification strategies. This can allow the mass spectrometer to be used solely for peptide identification. Laser-induced fluorescence detection provides a sensitive means for quantification of fluorescent analytes obtained from separation techniques such as capillary electrophoresis23-24 and liquid chromatography.25 At low fluorophore concentrations, quenching does not usually occur so that fluorescence intensity is directly proportional to concentration. This makes fluorescence detection applicable to microscale capillary electrophoresis or liquid chromatography. For fluorescence-based protein quantification, peptides should be derivatized with fluorescent reagents that have long absorbance and emission wavelength maximums, a detection limit in the subpicomolar range, and high water solubility. Additionally, a useful fluorescent label may react specifically with a subset of the peptide mixture to reduce sample complexity while maintaining a peptide population derived from the majority of the proteins in the sample. Theoretically, over 91.6% of the yeast proteins contain cysteines while only 9.3% of the tryptic (13) Desiderio, D. M.; K. M. Biomed. Mass Spectrom. 1983, 10, 471-9. (14) Kosaka, T.; T. T.; Nakamura, T. Anal. Chem. 2000, 72, 1179-85. (15) Yao, X.; F. A.; Ramirez, J.; Demirev, P. A.; Fenselau, C. Anal. Chem. 2001, 73, 2836-42. (16) Qin, J.; H. C.; Zhang, X. Rapid Commun. Mass Spectrom. 1998, 12, 20916. (17) Back, J. W.; N. V.; de Koning, L. J.; Muijsers, A. O.; Sixma, T. K.; de Koster, C. G.; de Jong, L. Anal. Chem. 2002, 74, 4417-22. (18) Wallis, T. P.; P. J.; Gorman, J. J. Protein Sci. 2001, 10, 2251-71. (19) Gygi, S. P.; Rist, B.; Gerber, S. A.; Turecek, F.; Gelb, M. H.; Aebersold, R. Nat. Biotechnol. 1999, 17, 994-999. (20) Smolka, M. B.; Zhou, H. L.; Purkayastha, S.; Aebersold, R. Anal. Biochem. 2001, 297, 25-31. (21) von Haller, P. D.; Yi, E.; Donohoe, S.; Vaughn, K.; Keller, A.; Nesvizhskii, A. I.; Eng, J.; Li, X. J.; Goodlett, D. R.; Aebersold, R.; Watts, J. D. Mol. Cell. Proteomics 2003, 2, 428-442. (22) Thompson, A.; Schafer, J.; Kuhn, K.; Kienle, S.; Schwarz, J.; Schmidt, G.; Neumann, T.; Hamon, C. Anal. Chem. 2003, 75, 1895-1904. (23) Cheng, Y. F.; Dovichi, N. J. Science 1988, 242, 562-564. (24) Craig, D. B.; Wong, J. C. Y.; Dovichi, N. J. Anal. Chem. 1996, 68, 697700. (25) Roach, M. C.; Harmony, M. D. Anal. Chem. 1987, 59, 411-415.

4496

Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

peptides from the yeast proteome have cysteine residues.26 This observation, coupled with the nucleophilic sulfhydryl group on its side chain, makes the cysteine residue an attractive target for fluorescent labeling. A previous study evaluated a fluorogenic thiol derivatization reagent, ammonium-7-fluoro-2,1,3-benzoxadiazole-4-sulfonate (SBDF), as a quantification tool for proteins.27 This reagent is watersoluble and has a fluorescent detection limit of 0.2 fmol for BSA, but the authors reported that MS detection of SBD-F-derivatized peptides is not optimal in the positive ion mode, presumably due to the negatively charged sulfonyl group.28 The same group synthesized positively charged fluorogenic thiol derivatization reagents, (4-(dimethylaminoethylaminosulfonyl)-7-chloro-2,1,3benzoxadiazole (DAABD-Cl) and 6-chloro-2,1,3-benzoxadiazole-4sulfonylaminoethyltrimethylammonium chloride (TAABD-Cl) for peptide fluorescence-based quantification and MS analysis. However, DAABD-Cl and TAABD-Cl have not been used to fractionate cysteine-containing peptides from complex peptide mixtures.28 As an alternative to these custom reagents, we investigated the utility of a common, inexpensive derivatization reagent called 1,5-I-AEDANS (5-[[2-[(iodoacetyl)amino]ethyl]amino]naphthalene1-sulfonic acid) for peptide quantification.29 Due to its long fluorescence lifetime and large Stokes shift, 1,5-I-AEDANS is often used for fluorescence polarization studies, rotational studies, and fluorescence resonance energy-transfer studies.30-31 Most importantly, 1,5-I-AEDANS label meets the previously mentioned criteria for a fluorescence-based quantification tool: (a) 1,5-I-AEDANS has distinct spectral properties from protein chromophores, with long absorbance and emission maximums of 336 and 490 nm, respectively,29 (b) the detection limit of AEDANS-labeled proteins is in the subpicomolar range,32 (c) the dye is highly soluble in aqueous solutions (to at least 0.05 M in phosphate buffer),29 and (d) 1,5I-AEDANS readily derivatizes cysteines,29 providing the means to selectively label a subset of tryptic peptide mixtures. In this study, we demonstrate that AEDANS-peptides produce readily interpretable positive ion MALDI-MS, ESI-MS, and CID tandem mass spectra with comparable sensitivity to unlabeled peptides. We also explore two AEDANS-peptide fractionation strategies. One AEDANS-peptide fractionation strategy employs Ga(III) immobilized metal affinity chromatography (IMAC)33 to recover AEDANS-labeled peptides and reduce the complexity of peptide mixtures. Using micro-LC-ESI-MS with an on-line fluorescence detector and a mass spectrometer, we generated fluorescence-based elution profiles and corresponding positive ion mass spectra of AEDANS-labeled peptides. In addition to the IMAC fractionation strategy, 1,5-I-AEDANS was coupled to an o-nitrobenzyl-based photocleavable resin to produce a solid phase that can label and isolate thiols and cysteine-containing peptides with a modified-AEDANS label (mAEDANS; 5-[(4-amino-4oxobutanoyl)2-[(iodoacetyl)amino]ethylamino]naphthalene-1(26) Wang, S.; Zhang, X.; Regnier, F. E. J. Chromatogr., A 2002, 949, 153-162. (27) Toriumi, C.; Imai, K. Anal. Chem. 2003, 75, 3725-3730. (28) Masuda, M.; Toriumi, C.; Santa, T.; Imai, K. Anal. Chem. 2004, 76, 728735. (29) Hudson, E. N.; Weber, G. Biochemistry 1973, 12, 4154-4161. (30) Jona, I.; Matko, J.; Martonosi, A. Biochim. Biophys. Acta 1990, 1028, 183199. (31) van Der Heide, U. A.; Orbons, B.; Gerritsen, H. C.; Y. K.; L. Eur. Biophys. J. 1992, 21, 263-272. (32) Gorman, J. J. Anal. Biochem. 1987, 26, 376-387. (33) Posewitz, M. C.; Tempst, P. Anal. Chem. 1999, 71, 2883-2892.

sulfonic acid). The o-nitrobenzyl-based photocleavable linker in this study has been used previously for solid-phase isotope tagging in proteomics applications.34 This fractionation protocol enriches cysteine-containing peptides more specifically than the IMAC strategy. MATERIALS AND METHODS Chemicals and Reagents. All peptides, bovine serum albumin (BSA), R-lactalbumin, 1,5-I-AEDANS, and other chemicals were purchased from Sigma-Aldrich (St. Louis, MO) unless otherwise specified. Sequencing grade trypsin was purchased from Roche Applied Sciences (Indianapolis, IN). Ga(III)-IMAC spin columns were obtained from Pierce Biotechnology, Inc. (Rockford, IL). Derivatization of Cysteine-Containing Peptides with 1,5I-AEDANS. The sequences for four peptides in these experiments are as follows: peptide A, CDPGYIGSR; peptide B; SVSEIQLC; peptide C, HCKFWW; and peptide D, RILAVERYLKDQQLLGIWGCSGK. Initial concentrations of the peptides were based on manufacturers’ estimation of peptide quantities. Peptides A-D were diluted in 50 mM ammonium bicarbonate (pH 8.5) and 0.2 mM TCEP to concentrations of 0.29, 0.27, 0.31, and 0.29 mM for peptides A-D, respectively. The peptides were incubated with the reducing agent for 30 min at 50 °C. The volumes of the peptide reaction mixtures were doubled and 1,5-I-AEDANS was added to a final concentration of 2.2 mM before incubation for 1 h. These reactions were carried out in triplicate for peptide C, and serial dilutions were made to generate a standard curve. IMAC Fractionation of Peptide Mixtures and Tryptic Digests. Peptide mixture 1 consisting of 11 µM peptide A, 10 µM peptide B, and 12 µM peptide C, and mixture 2 consisting of 11 µM peptide A, 10 µM peptide B, 24 µM peptide C were prepared in 2% acetic acid. Sixty microliters of each solution was loaded onto the Ga(III)-IMAC spin columns (Pierce Biotechnology, Inc.) for 15 min. The columns were washed two times with 100 µL of 0.1% acetic acid, followed by two 100-µL washes with 10% acetonitrile/0.1% acetic acid and one wash of 100 µL of water. The peptides were eluted with 100 µL of 0.3% ammonium hydroxide. For tryptic digests, 160 µg of R-lactalbumin and 60 µg of BSA were each denatured and reduced at 50 °C for 30 min in 25 µL of the following denaturing buffer: 8 M urea, 200 mM Tris, pH 8.5, 200 mM NaCl, 0.5 mM EDTA, and 2 mM TCEP. Ten microliters of 15 mM 1,5-I-AEDANS was added to each solution and the resultant mixture stirred for 30 min at room temperature. The samples were diluted with 200 µL of 50 mM ammonium carbonate (pH 8.5) and 2 µg of trypsin. The tryptic digest proceeded for 12 h at 37°C. A portion of each digest (25 µL) was mixed with 50 µL of 5% acetic acid, and 70 µL of each solution was loaded onto an IMAC spin column. The IMAC fractionation proceeded as described above. To remove the excess unreacted 1,5-I-AEDANS, the samples were incubated with 40 mg of thiopropyl Sepharose 6B that was preequilibrated with 50 mM ammonium bicarbonate (pH 8.5) and 0.25 mM TCEP. The samples reacted with the resin for 1 h with periodic agitation. The samples were removed from the resin and pooled with two subsequent washes: 100 µL of water followed by 100 µL of 10% acetonitrile. All samples were dried in a Rotovap and resuspended in 100 µL of 0.5% acetic acid for further analysis by LC-MS. (34) Zhou, H.; Ranish, J. A.; Watts, J. D.; Aebersold, R. Nat. Biotechnol. 2002, 19, 512-515.

LC-MS Detection of Labeled Peptides with On-Line Fluorescence Detection. Reversed-phase capillary HPLC was performed using a Zorbax C18 column (5 µm, 150 × 0.3 mm; Agilent Technologies, Palo Alto, CA) on a Michrom Paradigm MS4 HPLC fitted with a UV detector (Michrom Bioresources, Inc., Auburn, CA). Fluorescence chromatograms were collected using a Gilson 121 fluorometer fitted with a 0.6-µL flow cell (Gilson Co., Inc., Lewis Center, OH). The fluorometer excitation filter ranged from 310 to 410 nm with a maximum %T from 352 to 360 nm. The broad-band emission filter had a 50% T from 480 to 520 nm. For each LC-MS experiment, 20 µL of sample was injected onto the column. In most of the experiments, the on-line UV detector preceded the on-line fluorescence detector. For all LC-MS experiments, the flow rate was 12 µL/min. The LC solvent conditions for peptide C are as follows: the sample was loaded in 90% solvent A (0.1% formic acid) and 10% solvent B (0.1% formic acid in 99% acetonitrile) for 5 min. A 30-min linear gradient from 10 to 60% solvent B was followed by a 5-min wash in 80% solvent B and a 10-min reequilibration step in 10% solvent B. For LCMS experiments of peptide mixtures and tryptic digests, 20-µL aliquots of the peptides were loaded in 100% solvent A, followed by a 0-60% solvent B linear gradient for 75 min, a 5-min wash in 80% solvent B, and a 10-min reequilibration step. Flourescence, absorbance, and ESI-MS data were recorded using Masslynx Software (Waters Corp., Milford, MA). A Micromass Q-TOF was used to generate the ESI-MS data (Waters Corp.). MALDI-TOF MS Analysis. MALDI-TOF MS data were collected using a Perseptive BioSystem Voyager-DE STR mass spectrometer (Applied Biosystems, Foster City, CA). Peptides and AEDANS derivatives were crystallized with either R-cyano-4hydroxycinnamic acid or 2,5-dihydroxybenzoic acid matrix in the presence of 0.1% trifluoroacetic acid and 50% acetonitrile. CID of Labeled Peptides. Labeled peptides were directly infused into the Q-TOF for CID experiments. After parent ions were selected, collision energies were increased in increments of 5 V every 2 min in the presence of argon gas until the parent ion was completely dissociated (generally at 30-40 V). Generation of Photocleavable mAEDANS Resin. Five hundred milligrams of aminopropyl glass beads (170A, 200-400 mesh) were rinsed liberally with dimethylformamide (DMF) and coupled to an Fmoc-aminoethyl photocleavable linker (4-[2methoxy-4-(1-Fmoc-aminoethyl)-5 nitrophenoxy]butyric acid, Novabiochem) in the following manner: 49 mg of 1-hydroxybenzotriazole (Hobt; Aldrich) and 22.4 µL of 99% N,N′-diisopropylcarbodiimide (DIC) were mixed with 189 mg of the Fmoc-aminoethyl photolinker in 5 mL of DMF for 30 min. The solution was added to the resin and mixed overnight to form a carboxamide moiety between the resin and the aminoethyl photocleavable linker. The reaction proceeded until the Kaiser test showed no evidence of free amine. After the resin was washed with 15 mL of DMF, it was mixed with 2 mL of 40% (v/v) acetic anhydride in pyridine for 1 h in order to cap any remaining free amine. After washing the beads in 15 mL of DMF, the Fmoc-protecting group on the aminoethyl photocleavable linker was removed by mixing the beads with 5 mL of 20% (v/v) piperidine in DMF for 30 min. The presence of free amine was observed using the Kaiser test. A succinic acid group was coupled to the free amine by mixing 72.6 mg of succinic anhydride with the beads in 2 mL of 60% Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

4497

pyridine in DMF for 30 min until the Kaiser test was negative for the presence of free amine. The beads were washed with 20 mL of DMF. The beads were then mixed with 98 mg of Hobt and 112 µL of DIC for 30 min in 2 mL of 85:15 DMF.water. Then 314 mg of 1,5-I-AEDANS was diluted in 5 mL of 85:15 DMF/water and mixed with the resin. This slurry was agitated for 2-3 days. The beads were washed 10 times with 5 mL of 85:15 DMF/water solution, followed by three 5-mL ethanol washes. The beads were dried and stored in the dark at 4 °C. Solid-Phase mAEDANS Labeling and Isolation of a Thiol, Peptide, and Tryptic Digests. To derivatize a thiol with mAEDANS, 3 mg of the mAEDANS resin was rinsed 3 times with 1 mL of water. Then 100 µL of 5% β-mercaptoethanol in 20 mM ammonium bicarbonate (pH 8.5) was incubated with the resin for 30 min. The beads were washed 3 times with 1 mL of water. Then the beads were suspended in 100 µL of water and exposed to UV radiation with a Blak Ray UV lamp for 2 h with periodic agitation. To label peptide D, 10 µg of the peptide was incubated with 3 mg of mAEDANS resin for 30 min. The resin was rinsed 3 times with 1 mL of water, and the remaining unreacted mAEDANS resin was capped with 2% β-mercaptoethanol prior to exposure to UV light. To capture cysteine-containing peptides from tryptic digests, 190 µL of each of the digest mixtures was incubated with 10 mg of mAEDANS beads for 1 h. The beads were washed three times with 200 µL of 20 mM ammonum bicarbonate buffer (pH 8.5), once with 200 µL of 2 M NaCl, and three times with 200 µL of a 50:50 methanol/water solution. The beads were suspended in 100 µL of 20 mM ammonium carbonate buffer (pH 8.5) and exposed to UV light for 2 h. All cleaved products were dried down in a Rotovap and stored at -20 °C until further analysis. RESULTS AND DISCUSSION To determine whether AEDANS-labeled peptides can also be used for fluorescence-based peptide quantitation, a cysteinecontaining peptide was tagged with 1,5-I-AEDANS (structure 1),

and serial dilutions were made in order to perform multiple LCMS experiments using an on-line Gilson 121 fluorescence detector and a Micromass Q-TOF. The fluorometer and Q-TOF mass spectrometer had similar sensitivities. In addition, the Q-TOF positive ion mass spectra revealed that the detection sensitivity of AEDANS-labeled and unlabeled peptides were comparable. A standard curve was generated by fluorescence peak integration from eight samples (in triplicate) ranging from 2.7 to 46.5 pmol of AEDANS-HC*KFWW (where * denotes the site of the label). The fluorescence intensity is directly proportional to the AEDANSpeptide concentration (linear R2) 0.9949) and the LC-MS retention times of the highest and lowest concentrations tested are similar. Thus, LC-MS provides reproducible data, and fluorescence quenching does not occur at these concentrations. To determine whether product ion spectra of AEDANSpeptides are interpretable, CID experiments were performed with 4498 Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

Figure 1. CID of four AEDANS-peptides revealing interpretable product ion spectra. Product ion spectra of (A) AEDANS-peptide A (C*DPGYIGSR), (B) AEDANS-peptide B (SVSEIQLC*), (C) AEDANSpeptide C (HC*KFWW), and (D) AEDANS-peptide D (RILAVERYLKDQQLLGIWGC*SGK). A micromass Q-TOF was used to perform MS/ MS analysis of all peptides.

four AEDANS-labeled peptides using a Micromass Q-TOF (Figure 1). Three of the peptides contained under 10 amino acids (Figure 1A-C), while the fourth peptide was 23 amino acids long and represents a more complicated analysis of a multiply charged peptide with several “missed tryptic cleavage sites” (Figure 1D).

Scheme 1. IMAC Fractionation Protocol for AEDANS-Peptides

Two of the peptides contain a cysteine at either the N- or C-terminus, while the other two contain internal cysteine residues. The product ion spectra for peptides B (SVSEIQLC*) and C (HC*KFWW) are complete and reveal that the AEDANS label is intact and indeed attached to cysteine (the AEDANS label is denoted by a * in Figure 1 and corresponds to +306 m/z). The product ion spectrum for peptide A (C*DPGYIGSR, Figure 1A) contains most of the predicted y and b ions, with only the b3* and b8* ions missing from the spectrum. The missing b3* ion may be due to the poor fragmentation typically observed at glycine residues (located at position 4 in peptide A). Collision-induced dissociation of the +4 charge ion of peptide D (RILAVERYLKDQQLLGIWGC*SGK) presents a complex and incomplete product ion spectrum. However, several of the y and b ions are interpretable and reveal that the AEDANS label remains intact on the fragment ions that include the cysteine residue. The large mass of the label moves many of the y ions toward the higher m/z range in the spectrum, where there is an improved signalto-noise ratio (see the locations of y8* and y9* in Figure 1D). Overall, these experiments reveal the following: (1) despite the fact that the AEDANS-label contains a sulfonic acid moiety, AEDANS-peptides can be detected with good sensitivity using positive ion ESI-MS, and (2) AEDANS-peptides produce readily interpretable CID product ion spectra. Since AEDANS-peptides are fluorescent and provide suitable MS and product ion spectra, a relatively quick AEDANS-peptide fractionation protocol was developed using IMAC. Gallium(III)based IMAC has been used previously to fractionate phosphorylated peptides.33 Since AEDANS-peptides contain a sulfonic acid moiety, we reasoned that Ga(III)-IMAC could be used to bind and elute AEDANS-peptides as well. Using IMAC, the steps for AEDANS-peptide fractionation are described in Scheme 1. The protein solution is first reduced and labeled with 1,5-I-AEDANS and then digested with the appropriate endoprotease (i.e., trypsin). The peptide mixture is incubated with IMAC beads at a low pH. After multiple washes with dilute acetic acid followed by water, the bound peptides are eluted in an aqueous basic solution. To test whether IMAC is useful for AEDANS-peptide fractionation and quantitation studies, two standard peptide mixtures were labeled with 1,5-I-AEDANS and recovered using the IMAC fractionation protocol. The mixtures contained three AEDANSlabeled peptides: A, B, and C. Mixtures 1 and 2 contained identical concentrations of A and B, while mixture 2 contained twice the amount of AEDANS-peptide C. Each mixture was loaded onto an IMAC spin column in 2.5% acetic acid, washed several times with

Figure 2. Recovery of AEDANS-peptides after using IMAC. (A) Capillary reversed-phase HPLC chromatograms of AEDANS-peptide mixtures isolated, captured, and released from the IMAC resin. Mixture 1 and mixture 2 contained identical amounts of AEDANSlabeled peptides A and B. Mixture 2 contained twice the amount of AEDANS-peptide C. (B) Fluorescence chromatogram of tryptic AEDANS-peptides derived from BSA. Eight fluorescent peaks corresponded to the elution of AEDANS-labeled BSA peptides: 1, C*ASIQK; 2, SHC*IAEVEK; 3, LC*VLHEK; 4, GAC*LLPK; 5, RPC*FSALTPDETYVPK; 6, SLHTLFGDELC*K; 7, SLGKVGTRC+C+TKPESER (+ means only one of the cysteines is labeled) or LFTFHADIC*TLPDTEK; 8, MPC*TEDYLSLILNR, (C) Absorbance and corresponding fluorescence chromatogram of tryptic AEDANS-peptides derived from R-lactalbumin. On-line UV detection (330 nm) and the on-line fluorescence chromatogram revealed a total of five peaks corresponding to the elution of AEDANS-labeled R-lactalbumin peptides: 1, ALC*SEK; 2, IWC*K; 3, C*EVFR; 4, LDQWLC*EKL; 5, ALC*SEKLDQWLC*EKL.

0.2% acetic acid in the presence of 10% acetonitrile, washed with water, and eluted with 0.3% ammonium hydroxide at pH 11. Five microliters of the eluted product was diluted to 200 µL of 0.1% formic acid and loaded onto a reversed-phase C18 HPLC column with on-line fluorescence detection followed by ESI-MS analysis. The fluorescence-based elution profiles reveal that AEDANSAnalytical Chemistry, Vol. 77, No. 14, July 15, 2005

4499

Scheme 2. mAEDANS Solid-Phase Peptide-Labeling Strategy

peptides A-C were well resolved during the 0-60% acetonitrile gradient (Figure 2A). Fluorescent peak integration was performed, and peak ratios between mixtures 1 and 2 were tabulated. Fluorescence intensity ratios (mix 2/mix 1) of peptides A and B were 0.95 and 0.88, respectively, while the peptide C mix 2/mix 1 ratio was 2.10. These numbers correspond well with the input peak ratios, which were 1,1, and 2 for peptides A-C. Therefore, these results suggest that the IMAC protocol can be applied toward the accurate comparative analysis of AEDANS-peptide concentrations from multiple peptide samples. The AEDANS labeling strategy and IMAC fractionation protocol was tested on two proteins: BSA and R-lactalbumin. Each protein was reduced and labeled with excess 1,5-I-AEDANS. The labeled proteins were digested with trypsin. The digested products were loaded, washed, and eluted from an IMAC spin column. The post-IMAC sample was injected into a reversed-phase LC column with on-line fluorescence and ESI-MS detection. The post-IMAC fluorescence chromatogram revealed eight peaks that correspond to the mass tags of tryptic AEDANS-labeled BSA peptides (Figure 2B). Although the post-IMAC sample shared 20 total peaks with the pre-IMAC sample (including 12 peaks that corresponded to unlabeled and thus unobserved peptides by fluorescence detection), the post-IMAC sample contained 31 fewer MS peaks than the pre-IMAC sample. Therefore, the IMAC fractionation protocol reduced the sample complexity. Of these 31 unique pre-IMAC LC-MS peaks not observed post-IMAC, only 3 corresponded to cysteine-containing peptides. By contrast, eight (73%) AEDANSpeptides bound to and later eluted from the IMAC column. Figure 2C shows the fluorescence chromatogram of the postIMAC products derived from AEDANS labeling and tryptic digestion of R-lactalbumin. Both pre- and post-IMAC fluorescence chromatograms had five fluorescence peaks that corresponded to mass tags of AEDANS-labeled R-lactalbumin peptides. Once again, this demonstrates that AEDANS-peptides interact with the IMAC column. The dynamic range of detection and quantification can be enhanced by on-line UV absorbance detection at 330 nm. For example, Figure 2C shows an on-line absorbance chromatogram that corresponds to the fluorescence elution profile. In this case, peaks 2 and 3, which saturate the fluorescence detector, have well-defined peak maximums in the corresponding absorbance chromatogram. The use of both on-line fluorescence and absorbance detectors provide a larger dynamic range to quantify AEDANS-labeled peptides and eliminates the consideration of spurious peaks due to non-AEDANS impurities. Although the IMAC protocol shows promise as a rapid fractionation method for AEDANS peptides, unlabeled peptides also interact with the resin. For this reason, we developed another 4500

Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

Figure 3. Fluorescence and CID of mAEDANS products. (A) Fluorescence and selected ion chromatograms of a thiol labeled with mAEDANS. (B) The summed mass spectra corresponding to the chromatographic peak of the mAEDANS-labeled thiol. The data were collected in positive ion mode using a Micromass ESI-Q-TOF instrument. The base peak of the spectrum corresponds to the expected m/z of the mAEDANS-thiol. The expected structure of the mAEDANS-thiol is shown. (C) Product ion spectrum of mAEDANSpeptide D using ESI-MS with a Q-TOF. The m/z of the mAEDANS tag is +405.

more selective solid-phase approach for labeling and isolating cysteine-containing peptides (Scheme 2). Solid-phase peptide synthesis chemistry was used to couple 1,5-I-AEDANS to glass beads containing a photocleavable linker, which forms the mAEDANS resin (structure 2).

Figure 4. Tryptic mAEDANS-peptides captured, labeled, and released from the mAEDANS resin. (A) MALDI-TOF MS of R-lactalbumin tryptic peptides. (B) MALDI-TOF MS mAEDANS-R-lactalbumin photocleaved products. The labeled peaks correspond to masses (m/z) of mAEDANSlabeled R-lactalbumin tryptic peptides. (C) MALDI-TOF MS of BSA tryptic peptides. (D) MALDI-TOF MS of mAEDANS-BSA photocleaved products. The labeled peaks correspond to masses (m/z) of mAEDANS-labeled BSA tryptic peptides.

In this protocol, the proteolyzed mixture is reduced and cysteinylpeptides can be attached to the mAEDANS resin in a basic solution. After washing away impurities, the mAEDANS-peptides can be cleaved from the beads using UV radiation. To test the resin for its ability to capture and label a thiol with mAEDANS, β-mercaptoethanol was incubated with the mAEDANS resin in ammonium bicarbonate buffer (pH 8.5) for 15 min. The resin was washed liberally with buffer, a NaCl solution, and water followed by exposure to UV radiation for 2 h to cleave the product. Initially, the product was observed by MALDI-TOF MS in the negative ion mode and corresponded to the expected m/z of 482. Subsequently, the product was injected onto a reversed-phase HPLC column with on-line fluorescence and MS detection using the ESI-Q-TOF in positive ion mode. The chromatogram in Figure 3A reveals that the major fluorescence peak corresponds to the selected ion chromatogram of the expected [M + H]+ ion of the mAEDANS-labeled thiol. Figure 3B is the summed mass spectra derived from the peak in the selected ion chromatogram. To determine whether the resin could be used to label a long peptide with mAEDANS, peptide D was incubated with the resin in ammonium carbonate buffer for 1 h, washed, and exposed to UV radiation as described in Scheme 2. The cleaved product was first analyzed by MALDI-TOF MS and revealed the expected m/z of the mAEDANS-peptide in positive ion mode. Subsequently, ESIMS using a Q-TOF mass spectrometer was performed on mAEDANS-peptide D, and CID of the [M + 3H]3+ ion produced readily interpretable product ions (Figure 3C). Despite the peptide complexity, most of the b ions and eight of the y ions were

identified. In this experiment, the mAEDANS label appears to be intact in the product ions that included the cysteine (represented by a *, corresponded to +405 m/z). To determine whether the mAEDANS resin can be used to label tryptic digest peptide mixtures, R-lactalbumin and BSA were reduced and digested with trypsin and captured by the mAEDANS resin (Scheme 2). The initial digested samples (Figure 4A and C) and the cleaved/eluted products (Figure 4B and D) were analyzed by MALDI-TOF MS. Several mass tags corresponding to mAEDANS-labeled peptides and Na+ adducts of mAEDANS peptides were identified. Overall, these experiments reveal that the mAEDANS resin can be used to enrich and label cysteinylpeptides with a fluorescent tag. The mAEDANS products can lead to interpretable product ion spectra with CID, demonstrating that mAEDANS can be used for peptide sequence identification. CONCLUSIONS The 1,5-I-AEDANS molecule is not only inexpensive and commercially available, but the fluorescence intensity of an AEDANS-peptide is also directly proportional to the peptide concentration. Thus, 1,5-I-AEDANS can be used for fluorescencebased peptide quantification. The AEDANS-peptides produce CID product ion spectra with the label intact. So the experimenter can easily interpret the product ion spectra manually or with an automated de novo sequencing program once the cysteine modification is included in the search parameters. Two AEDANS-peptide fractionation studies were developed. The first study demonstrated that AEDANS-peptides can be Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

4501

fractionated using Ga(III)-IMAC. In this study, we show that IMAC fractionation of two peptide mixtures results in relative yields of AEDANS-peptides that accurately reflect the input. Additionally, on-line fluorescence chromatograms of AEDANS-peptides during an LC-MS experiment can assist the experimenter in discovering low-abundance AEDANS-peptides in the samples. For example, in Figure 4C, the BSA-derived peptide elution profiles generally were not observed from the total ion chromatograms. Many of these ions were observed only after summing the spectra that corresponded to the fluorescence peaks. Additionally, an on-line UV detector set at 330 nm can be used to complement the fluorescence results, increasing the dynamic range of AEDANSpeptide quantification. A second AEDANS-peptide fractionation strategy employed a modified AEDANS (mAEDANS) molecule linked to a photocleavable resin. The mAEDANS resin was developed through standard solid-phase peptide synthesis chemistry. With this resin, we labeled a thiol with the mAEDANS compound and further demonstrated that mAEDANS can label peptides and produce interpretable CID product ion spectra with the fluorescence mAEDANS tag intact. To increase the versatility of mAEDANS, the experimenter can also create a mAEDANS isotope coded label by introducing stable isotopes through the succinamic acid moiety during synthesis. Despite the advantages of this solid-phase labeling/isolation strategy, the major drawbacks to using a photocleavable linker with 1,5-I-AEDANS are as follows: (1) the photocleavable linker is relatively expensive, and (2) the experimenter has to cap the compound with a thiol before cleaving the labeled products

because the iodoacetamide moiety causes 1,5-I-AEDANS to degrade when exposed to light.29 These limitations make the task of removing unreacted mAEDANS products prior to LC-MS more difficult. Excess mAEDANS can produce a high fluorescence background during LC-MS experiments. Alternatively, acid-labile linkers, such as a Rink Amide Linker, could be used to create a solid-phase mAEDANS resin instead of the photocleavable linker. This would allow the excess unreacted mAEDANS to remain intact upon elution of mAEDANS products from the resin. The iodoacetamide moiety on the eluted mAEDANS molecules could then be exploited to selectively remove excess mAEDANS from the labeled-peptides prior to LC-MS experiments. In support of this, a Rink Amide Linker has been used previously to create iodo-containing resins for labeling cysteinylpeptides during proteomic analysis.35 This study shows that 1,5-I-AEDANS provides a simple fluorescence-based peptide quantification tool that is compatible with standard MS techniques. In addition, two fractionation protocols were developed to simplify peptide mixtures that are labeled with this compound, enhancing the versatility of 1,5-IAEDANS as a proteomics tool.

(35) Zhang, L.; Guo, Y.-L.; Liu, H.-Q. J. Mass Spectrom. 2004, 39, 447-457.

AC050247K

4502

Analytical Chemistry, Vol. 77, No. 14, July 15, 2005

ACKNOWLEDGMENT The authors thank M. A. Scialdone, Q. T. Shang, B. AmiriEliasi, and J. M. Ginter for useful discussions. This research was supported in part by NSF grants CHE9634238 and DBI0096578. Received for review February 9, 2005. Accepted April 19, 2005.