Anal. Chem. 2002, 74, 5443-5449
Articles
Probing Lysine Acetylation with a Modification-Specific Marker Ion Using High-Performance Liquid Chromatography/ Electrospray-Mass Spectrometry with Collision-Induced Dissociation Jin Young Kim,†,‡ Kyoung Wook Kim,† Ho Jeong Kwon,§ Dai Woon Lee,‡ and Jong Shin Yoo*,†
Proteome Analysis Team, Korea Basic Science Institute, 305-333, Daejeon, Korea, Department of Bioscience and Biotechnology, Sejong University, 143-747, Seoul, Korea, and Department of Chemistry, Yonsei University, 120-749, Seoul, Korea
Posttranslational acetylation of proteins regulates many diverse functions, including DNA recognition, proteinprotein interaction, and protein stability. The identification of enzymes that regulate protein acetylation has revealed broader use of this modification than was previously suspected. In this study, we describe a method for identifying protein acetylation at lysine residues by analysis of digested protein using HPLC/ESI-MS with a new modification-specific marker ion. Collision-induced dissociation with capillary or nano-LC/ESI-TOF-MS was used to obtain a fragment ion useful as a marker for acetylated lysine. Although the acetylated lysine immonium ion at m/z 143.1 has been used as a marker ion for detecting acetylated lysine, it can be confused with internal fragment ion in some peptides, producing false positive results. We have found a novel marker ion at m/z 126.1, which is a further fragment ion induced by the loss of NH3 from the acetylated lysine immonium ions at m/z 143.1. This novel marker ion was found to be more specific and ∼9 times more sensitive than the immonium ion at m/z 143.1. In addition, no interfering ions for acetylated peptides were found in the extracted ion chromatogram at m/z 126.1. The utility of this method was demonstrated with acetylated cytochrome c as a model compound. After the modification was probed by the new marker ion, the acetylated lysine site was determined by the CID-MS spectrum. This method was applied to identify histone H4 acetylation in HeLa cells treated with trichostatin A. Three protein bands separated by acid-urea-Triton gel elec10.1021/ac0256080 CCC: $22.00 Published on Web 09/28/2002
© 2002 American Chemical Society
trophoresis were confirmed as tetra, tri, and diacetylated histone H4 at lysines 5, 8, 12, and 16. This method may be useful for assaying for lysine acetylation, which is an important regulatory process for a range of biological functions. Genes determine the sequence of a protein. Immediately following gene translation, the resulting protein is composed of the 20 common amino acids. A wide variety of enzymes subsequently modify many proteins in some manner. These posttranslational modifications include alterations of specific amino acid side chains. Because many biological processes regulating protein function are controlled by such modifications, their analysis is an essential step in the mechanistic investigation of biological pathways.1 In vivo, acetylation is an important covalent protein modification. It occurs at free amino groups at both the amino terminus and -amino group of lysine residues.2 Posttranslational modification of lysine residues may alter the solubility and conformation of a protein by changing the net charge of the molecule. In addition, acetylation can modify DNA recognition, protein stability, and protein interaction.3 Acetylation has also been reported in a * Corresponding author. Tel: 82-42-865-3432. E-mail:
[email protected]. † Korea Basic Science Institute. ‡ Yonsei University. § Sejong University. (1) Han, K. K.; Martinage, A. Int. J. Biochem. 1992, 24, 19-28. (2) Allfrey, V. G.; Di Paola, E. A.; Sterner, R. Methods in Enzymology; Academic Press: London, 1984; Vol. 107, pp 224-240. (3) Kouzarides, T. EMBO J. 2000, 19, 1176-1179.
Analytical Chemistry, Vol. 74, No. 21, November 1, 2002 5443
growing variety of proteins including histones, DNA-binding transcription factors, acetylases, nuclear import factors, and R-tubulin.4-6 Furthermore, it has been well documented that many acetylated proteins, including histones, are involved in transcriptional regulation.7-9 Various analytical techniques are available for the characterization of covalent protein modifications, such as enzymatic or chemical cleavage, isoelectric focusing, affinity chromatography, radiolabeling, specified derivatization protocols, and mass spectrometry.10-12 Protein sequencing by Edman degradation is not practical for detecting protein modifications because of their instability during this type of analysis. Mass spectrometry has become a useful tool for the characterization of peptides and proteins because it can determine accurate molecular weights and provide amino acid sequence information by tandem mass spectrometry.13-16 One of the most important applications of mass spectrometry in peptide sequencing is the identification of posttranslational modifications.17-19 Because mass spectrometry provides direct evidence of structural change, it is perhaps the best technique available to identify protein modification. Tandem mass spectrometry analysis of peptides can be performed by either collision-induced dissociation (CID) in a collision cell or in-source fragmentation in an electrospray ionization (ESI) source. The former requires a tandem mass spectrometer and the latter an additional tool, such as high-performance liquid chromatography (HPLC), to separate compounds in a mixed sample. The combination of HPLC with ESI mass spectrometry is a well-established method for the identification and characterization of peptides and proteins.20 HPLC/MS yields mass spectrometric data that provide structural information, while maintaining the separation obtained from the chromatographic analysis. The use of on-line HPLC/MS permits the identification of modified peptides from peptide digest mixtures by screening characteristic fragment ions related to modification. Carr et al. introduced the use of specific marker ions for the detection of special modifications such (4) Thirne, A. W.; Kmiciek, D.; Mitchelson, K.; Sautiere, P.; Crane-Robinson, C. Eur. J. Biochem. 1990, 710-713. (5) Turner, B.; J Cell Sci. 1991, 99, 13-20. (6) Smith, B. F.; Schwartz, B. L.; Randall, L. L.; Smith, R. D. Protein Sci. 1996, 5, 488-494. (7) Lin, P. P.; Barry, R. C.; Smith, D. L.; Smith, J. B. Protein Sci. 1998, 7, 14511457. (8) L’Hernault, S. W.; Rosenbaum, J. L. Biochemistry 1985, 24, 473-478. (9) Bayle, J. H.; Crabtree, G. R. Chem. Biol. 1997, 4, 885-888. (10) Imahori, K., Sakiyama, F., Eds. Methods in Protein Sequence Analysis; Plenum Press: New York, 1993. (11) Han, K. K.; Martinage, A. Int. J. Biochem. 1993, 25, 957-970. (12) Gianazza, E. J. Chromatogr., A 1995, 705, 67-87. (13) Winston, R. L.; Fitzgerald, C. M. Mass Spectrom. Rev. 1997, 16, 165-179. (14) Wilm, M.; Shevchenko, A.; Houthaeve, T.; Breit, S.; Schwegerer, L.; Fotsis, T.; Mann, M. Nature 1996, 379, 466-469. (15) Loo, J.; Loo, R. R. O. In Electrospray Ionization Mass Spectrometry; Cole, R. B., Ed.; John Wiley & Sons: New York, 1997; pp 385-419. (16) Hunt, D. F.; Shabanowitz, J.; Yates, J. R.; Griffin, P. R. In Mass Spectrotmetry of Biological Materials; McEwen, C. N., Larsen, B. S., Eds.; Marcel Dekker: New York, 1990; pp 169-195. (17) Carr, S. A.; Roberts, G. D.; Hemling, M. E. In Mass Spectrotmetry of Biological Materials; McEwen, C. N., Larsen, B. S., Eds.; Marcel Dekker: New York, 1990; pp 87-136. (18) Wilm, M.; Neubauer, G.; Mann, M. Anal. Chem. 1996, 68, 527-533. (19) Kouach, M.; Belaiche, D.; Jaquinod, M.; Couppez, M.; Kmiecik, D.; Ricart, G.; Van Dorsselaer, A.; Sautiere, P.; Briand, G. Biol. Mass Spectrom. 1994, 23, 283-294. (20) Fligge, T. A.; Bruns, K.; Przybylski, M. J. Chromatogr., B 1998, 706, 91100.
5444
Analytical Chemistry, Vol. 74, No. 21, November 1, 2002
as phosphorylation, glycosylation, and sulfonation.21,22 More recently, Jedrzejewski and Lehmann reported the detection of multiple modified peptides using programmable skimmer CID acquisition. Fragment ions were observed at low mass under single ion monitoring conditions and subsequently molecule ions at high-mass range in a single scan cycle.23 Borchers et al. selected the acetylated lysine immonium ion at m/z 143.1 as a specific marker for detecting acetylated lysine.24 However, in our study, the ion at m/z 143.1 was confused with other fragment ions produced by internal cleavage in some peptides. Therefore, it was not sufficient to probe acetylated lysine by the ion at m/z 143.1. In this study, we report a new specific marker ion for acetylated lysine with higher selectivity and sensitivity than the ion at m/z 143.1 and describe a way to identify lysine acetylation by the analysis of digested protein using HPLC/ESI-MS. This method was demonstrated in an analysis of acetylated cytochrome c, a model acetylated protein. The method was also applied to identify acetylation in the human histone protein, histone 4 (H4) from HeLa cells. H4 is one of several histones associated noncovalently with DNA in a typical eukaryotic nucleus.25 Higher levels of H4 acetylation have long been associated with the unwinding of chromatin for transcription. We identified acetylated H4 proteins, each with a different number of acetylations and different acetylation sites, from human HeLa cells treated with trichostatin A (TSA). MATERIALS AND METHODS HPLC-grade solvents were obtained from B&J. Horse heart cytochrome c and partially acetylated horse heart cytochrome c were obtained from Sigma. Modified sequencing-grade trypsin was obtained from Boehringer Mannheim Biochemicals. Acetic acid was obtained from Sigma. Mass Spectrometry Analysis. Capillary-LC/MS. Analyses of cytochrome c and acetylated cytochrome c were performed by capillary-LC/MS. A C18 capillary column (i.d. 300 µm, length 150 mm, particle size 5 µm; LC Packings) was used to separate peptide digests. Chromatography was performed using an LC-10AD (Shimazu) pump system. A precolumn splitter (Accurate; LC Packings) established a flow rate of 3 µL/min through the column from a delivery flow rate of 0.2 mL/min. The mobile phase used was 5% acetonitrile in 0.1% aqueous acetic acid (solvent A) and 95% acetonitrile in 0.1% aqueous acetic acid (solvent B). A solvent gradient of 5% B to 50% B over 100 min was used. A Famos autosampler (LC Packings) was used to produce an injection volume of 1 µL. Electrospray mass spectra were acquired on a Mariner mass spectrometer (Perkin-Elmer) equipped with an electrospray source. Solution was sprayed at a potential of ∼3.5 kV. To confirm molecular weight of digested peptides, in the first analysis, capillary-LC/ESI-MS was performed at a nozzle voltage of 125 V. Then, in the second, the nozzle voltage was increased to 225 V to fragment peptides, and CID-MS spectra by in-source fragmentation were obtained. (21) Huddleston, M. J.; Bean, M. F.; Carr, S. A. Anal. Chem. 1993, 65, 877884. (22) Annan, R. S.; Carr, S. A. J. Protein Chem. 1997, 16, 391-402. (23) Jedrzejewski, P. T.; Lehmann, W. D. Anal, Chem. 1997, 69, 294-301. (24) Borchers, C.; Parker, C. E.; Deterding, L. J.; Tomer, K. B. J. Chormatogr., A 1999, 854, 119-130. (25) Elglin, S. C. R., Ed. Chromatin Structure and Gene Expression; IRL Press: New York, 1995.
Nano-LC/MS. H4 analysis was performed using a Nano-LC/ MS system consisting of an Ultimate HPLC system (LC Packings) for nano-LC and a Q-TOF2 mass spectrometer (Micromass) equipped with a nano-ESI source. A C18 nanocolumn (i.d. 75 µm, length 150 mm, particle size 5 µm; LC Packings) was used. The flow rate through the column was 200 nL/min. A wateracetonitrile gradient was employed with both mobile phases containing 0.1% acetic acid. The gradient used was 0%-50% acetonitrile over 200 min. The injection volume was 1 µL. In the nanoelectrospray ionization source, the end of the capillary tubing from the nano-LC column was connected to the emitter with pico-tip silica tubing (i.d. 8 µm; New Objectives) by a stainless steel union, with a PEEK sleeve for coupling the nanospray with the on-line nano-LC. The applied voltage to the union to produce an electrospray was 2 kV, and cone voltage was 30 V. In the first nano-LC/ESI-MS analysis, the molecular weight of each digested peptide was determined from the MS spectrum obtained at collision energy of 10 eV. In the second, collision energy was increased to 30 eV to obtain fragment ions in MS mode without precursor selection and CID-MS spectra were obtained. Argon was introduced as a collision gas at a pressure of 10 psi. Sample Preparation. Cytochrome c and partially acetylated cytochrome c were prepared at 20 pmol/µL and digested with trypsin overnight at 37 °C in 50 mM NH4HCO3 at pH 8.5 (protein: enzyme ) 50:1, w/w). HeLa cells were treated with 200 ng/mL TSA for 18 h at 37 °C, and histones were extracted as described by Kwon et al.26 Histones, including H4 and its acetylated forms, were separated by acid-urea-Triton (AUT) gel electrophoresis, because of its utility for resolving basic proteins. The stained protein bands were excised from the gel and digested with trypsin as previously described.14,27 After being washed with 10 mM NH4HCO3 and acetonitrile, gel pieces were swollen in digestion buffer containing 50 mM NH4HCO3, 5 mM CaCl2, and 12.5 ng/µL trypsin and incubated at 37 °C for 12-16 h. The peptides were recovered by two extractions with 50 mM NH4HCO3 and acetonitrile. The resulting peptide extracts were pooled and lyophilized in a vacuum centrifuge and stored at -20 °C. RESULTS AND DISCUSSION Cytochrome c is a 12 360-Da protein with a single polypeptide chain of 104 amino acids that includes 19 lysine residues and a covalently attached heme group. Acetylated cytochrome c is heterogeneously acetylated at ∼60% of its lysine residues. ESI was used to elucidate the difference in the masses of cytochrome c and acetylated cytochrome c. The acetylated cytochrome c showed a 42 u mass difference between partially acetylated mixtures. To study posttranslational modifications in detail by mass spectrometry, the protein must first be enzymatically cleaved to produce peptides. To identify acetylated residues, acetylated cytochrome c was digested with trypsin and the resulting peptides were analyzed by capillary LC/ESI-TOF-MS. Due to the heterogeneity of acetylated cytochrome c, many of the digested peptides were present in the total ion chromatogram (TIC). The molecular weight of each component could be identified by the deconvolution (26) Kwon, H. J.; Owa, T.; Hassig, C. A.; Shimada, J.; Schreiber, S. L. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 3356-3361. (27) Jeno, P.; Mini, T.; Moes, S.; Hintermann, E.; Horst, M. Anal. Biochem. 1995, 224, 75-82.
Figure 1. Capillary-LC/ESI-CID-MS spectra of peptide MIFGIK (A) and acetylated peptide MIFGIKAcK (B) from trypsin digests of acetylated cytochrome c. KAc, acetylated lysine.
Figure 2. Total ion chromatogram of acetylated cytochrome c trypsin digests (A) and extracted ion chromatograms of the ions at m/z 143.1 (B) and 126.1 (C).
of a series of multiply charged molecular ions. Then, CID-MS experiments were performed to locate and sequence the peptides containing acetylated lysine residues. The instrument (nozzle) voltage was increased to a value where fragmentation could be Analytical Chemistry, Vol. 74, No. 21, November 1, 2002
5445
Table 1. Peak Identifier, Theoretical (cal), and Experimental (exp) Mass of the Trypsin Digests of Acetylated Cytochrome c M+H peak
cal
exp
trypsin digests residue
sequence
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
374.2 317.2 575.3 745.4 604.3 804.5 848.5 1475.8 1640.8 1754.8 1924.9 949.6 1119.7 851.4 1021.5 1547.9 2552.3 2382.2 2225.9 1650.9 2883.3 3053.4 3417.6 4074.9 4245.0 1918.8 4472.2 2249.2 2576.2 2746.3 2973.4 2123.0 2293.1 3278.7
374.4 317.3 575.5 745.6 604.5 804.7 848.7 1476.2 1641.2 1755.3 1925.4 949.8 1120.0 851.4 1021.8 1548.3 2552.9 2382.8 2225.5 1651.3 2883.1 3053.2 3417.4 4074.8 4245.4 1918.3 4472.4 2249.8 2575.8 2745.9 2973.1 2123.4 2293.6 3279.1
6-8 87-88 88-91 87-91 100-104 8-13 73-79 26-38 39-53 40-55 39-55 80-87 80-88 74-80 73-80 80-91 39-60 40-60 14-27 74-87 9-27 8-27 14-38 9-38 8-38 14-25 6-38 74-91 9-25 8-25 6-25 56-72 56-73 54-79
GKAcK KAcK KAcTER KAcKAcTER KAcATNE KAcIFVQK KAcYIPGTK HKAcTGPNLHGLFGR KAcTGQAPGFTYTDANK TGQAPGFTYTDANKAcNK KAcTGQAPGFTYTDANKAcNK MIFAGIKAcK MIFAGIKAcKAcK YIPGTKAcM KAcYIPGTKAcM MIFAGIKAcKAcKAcTER KAcTGQAPGFTYTDANKAcNKAcGITWK TGQAPGFTYTDANKAcNKAcGITWK CAQCHTVEKAcGGKAcHK+heme YIPGTKAcMIFAGIKAcK IFVQKAcCAQCHTVEKAcGGKAcHK+heme KAcIFVQKAcCAQCHTVEKAcGGKAcHK+heme CAQCHTVEKAcGGKAcHKAcTGPNLHGLFGR+heme IFVQKAcCAQCHTVEKAcGGKAcHKAcTGPNLHGLFGR+heme KAcIFVQKAcCAQCHTVEKAcGGKAcHKAcTGPNLHGLFGR+heme CAQCHTVEKAcGGK+heme KAcKAcIFVQKAcCAQCHTVEKAcGGKAcHKAcTGPNLHGLFGR+heme YIPGTKAcMIFAGIKAcKAcKAcTER IFVQKAc CAQCHTVEKAcGGK+heme KAcIFVQKAcCAQCHTVEKAcGGK+heme GKAcKAcIFVQKAcCAQCHTVEKAcGGK+heme GITWKAcEETLMEYLENPK GITWKAcEEYLMEYLENPKAcK NKAcGITWKAcEETLMEYLENPKAcKAcYIPGTK
induced by in-source fragmentation occurring in the electrospray source. Two of the resulting CID-MS spectra are shown in Figure 1 (MH+ 779.7 is shown in Figure 1A, and MH+ 949.8 in Figure 1B). The CID-MS spectra of (A) and (B) showed a similar pattern, with the same mass differences at corresponding y ions. However, the mass difference between y1 and y2 is 170 Da in (B), suggesting that the second amino acid from the C-terminus is the acetylated lysine with a molecular weight of 170 and a mass 42 u higher than the 128-Da lysine. The peptide with MH+ 779.7 corresponds to the peptide MIFAGIK (80-86) of cytochrome c, and the peptide with MH+ 949.8 corresponds to the acetylated peptide MIFAGIKAcK (80-87) with one acetylated lysine (KAc), where trypsin cannot recognize the acetylated lysine. In this way, peptides containing acetylated lysine can be identified from their CID-MS spectra. Detection of the modified peptides from a complex mixture by the above method is difficult and time-consuming because all peptides should be sequenced. To identify peptides containing an acetylated lysine easily and selectively, we applied an approach that detects a modification-specific marker ion produced by insource fragmentation or CID during the course of HPLC/ESIMS analysis. The protein digests are separated by HPLC, and the modified peptides are located by examining marker ion traces in an acquired CID-MS spectrum. Once the modified peptides have been located, they can be sequenced by their CID-MS spectra. 5446 Analytical Chemistry, Vol. 74, No. 21, November 1, 2002
Upon collision, fragmentation takes place along the peptide backbone. Peptides containing acetylated lysine residues produce the acetylated lysine immonium ion at m/z 143.1, 42 u higher than the unmodified lysine immonium ion at m/z 101.1.The resulting TIC (A) and the extracted ion chromatogram (EIC) of the ion at m/z 143.1 (B) for the trypsin digests of acetylated cytochrome c are shown in Figure 2. Although the EIC of the ion at m/z 143.1 shows several peaks, the number is much smaller than expected. The peptides were sequenced by their CID-MS spectra, but only a few were identified as acetylated lysine-containing peptides. Unexpectedly, peptides corresponding to peaks G, I, and J were not acetylated, despite showing strong peaks in the EIC of the ion at m/z 143.1. The nonacetylated peptides were identified as follows: G, GITWK (56-60); I, TGPNLHGLFG (28-38); and J, MIFAGIK (80-86). The three peaks present in the EIC of the ion at m/z 143.1 not attributed to the acetylated lysine immonium ion but to internal cleavage fragment ion appear to have resulted from peptides containing GI or GL amino acid sequences. The ion at m/z 143.1 is the dipeptide immonium ion of GI or GL, which has lost a carbonyl group from their acylium ion by internal fragmentation. Therefore, it appears that the ion at m/z 143.1 is neither specific nor sensitive enough for use as a modificationspecific marker for acetylated lysine detection. However, a novel effective marker ion for probing acetylated lysine was identified. In the CID-MS spectrum of the peptide
Figure 3. Capillary-LC/ESI-CID-MS spectrum of acetylated peptide peak 16 (MIFAGIKAcKAcKAcTER) from the trypsin digests of acetylated cytochrome c KAc, acetylated lysine.
Figure 5. Commassie-stained AUT gel analysis of histones from HeLa cells. HeLa cells treated without (control) and with TSA.
Figure 4. Total ion chromatogram of cytochrome c trypsin digests (A) and extracted ion chromatograms of the ion at m/z 143.1 (B) and 126.1 (C).
MIFAGIKAcK (Figure 1B), we observed an ion at m/z 126.1 that was not seen in the peptide MIFAGIK (Figure 1A). Because the two peptides differ by only one residue, acetylated lysine, we assumed that the ion at m/z 126.1 was the result of lysine acetylation, where the ion contains the modified structure of acetylated -amine. The ion at m/z 126.1 appears to result from the acetylated lysine immonium ion at m/z 143.1 due to a loss of NH3. In FAB-MS spectrum at low-energy CID, Dookeran et al. also observed the elimination of ammonia from the immonium ion of aliphatic amino acids by further fragmentation.28 The EIC of the ion at m/z 126.1 is shown in Figure 2C. Many more peaks, including those of acetylated peptides previously
identified in the EIC of the ion at m/z 143.1, were identified. However peaks G, I, and J were not seen, consistent with a lack of acetylated lysine residues. Most of the detected peptides produced a sufficient signal to be positively identified by the ion at m/z 126.1. The peptide at peak 16 was sequenced, and the acetylation site was identified by its CID-MS spectrum (Figure 3). As expected, it contained acetylated lysine residues. Interestingly, the peptides at peak 14 (Figure 2C) with MH+ 851.4 and peak 15 with MH+ 1021.8, which showed the ion at m/z 126.1, did not correspond to any peptide expected from the trypsin digests of acetylated cytochrome c. Analysis of their CID-MS spectra identified them as peptides YIPGTKAcM (74-80) and KAcYIPGTKAcM (73-80), which resulted from acetylated cytochrome c via an unexpected cleavage. We identified all the peaks detected by the ion at m/z 126.1 by molecular weight and by sequence analysis by CID-MS spectra (Table 1). All the identified peptides that showed a peak in the EIC of the ion at m/z 126.1 contained acetylated lysine residues. There were no interfering peaks for detecting the peptides containing acetylated lysine. These results suggest that the ion at m/z 126.1 is a marker ion that indicates acetylated lysine. Furthermore, the ion at m/z 126.1 is ∼9 times more sensitive than the acetylated lysine immonium ion at m/z 143.1 in the intensity of EIC. To evaluate selectivity in probing for acetylated lysine, nonacetylated cytochrome c was analyzed in the same manner. The resultant TIC and EICs for the two ions at m/z 143.1 and 126.1 are shown in Figure 4. The acetylation-selective trace for cytochrome c shows no peaks to consider for the ion at m/z 126.1. Although, there are a few peaks at very low intensity, it appeared that they are spurious on analysis of their CID-MS spectra where the ion at m/z 126.1 was observed at background level. In a comparison of TIC and the EIC for the ion at m/z 126.1, two chromatograms show a similar pattern. We think the peaks in the trace of the m/z 126.1 ion were attributed to the increase of total ions entering into the MS analyzer. However, there are three (28) Dookeran, N. N.; Yalcin, T.; Harrison, G. J. Mass Spectrom. 1996, 31, 500508.
Analytical Chemistry, Vol. 74, No. 21, November 1, 2002
5447
Figure 6. Total and extracted ion chromatograms of the marker ion at m/z 126.1 of the trypsin digests of band A (A), band B (B). and band C (C) shown in the AUT gel of HeLa cells treated with TSA Peak ID (mark:MH+:sequence): 1:1325.9:DNIQGITKPAIR, 2:1180.8: ISGLIYEETR, 3:989.7:VFLENVIR, 4:1326.7:TVTAMoxiDVVYALK, and 5:714.4:TLYGFGG. Moxi, oxidized methionine.
peaks for the ion at m/z 143.1, which are the same peaks shown in trypsin digests of acetylated cytochrome c (Figure 2B). They
correspond to the peptides that produced the dipeptide immonium of GI or GL at m/z 143.1 by internal fragmentation. The peptides have already been confirmed in the analysis of acetylated cytochrome c. In a comparison of the two EICs of the acetylated lysine immonium ion at m/z 143.1 and the new marker ion at m/z 126.1 used for detection of peptides containing acetylated lysine, the traces showed a high degree of peak center coincidence. However, the ion at m/z 143.1 was much less sensitive than the ion at m/z 126.1 and could be produced from the internal fragmentation of some peptides containing GI or GL amino acid sequences causing false results. Therefore, it may be insufficient to detect acetylated lysine selectively with the ion at m/z 143.1. In contrast, the ion at m/z 126.1 showed a good selectivity and sensitivity for all modified peptides containing acetylated lysine, as indicated by the quality of the EICs. The ion at m/z 126.1 is the first specific marker ion to detect peptides containing acetylated lysine The above method was applied to identify the acetylation of the human protein H4 from HeLa cells grown both with and without TSA treatment. TSA, an inhibitor of histone deacetylase, is known to induce the acetylation of histone.26 Identification of acetylation in histone proteins is important for studying the effect of acetylation on chromatin structure during transcription. Histone proteins were separated by AUT gel electrophoresis, and images of gels containing intact histones including H4 are shown in Figure 5. Three new bands, A-C are seen in lane 2, which was HeLa cells treated with TSA. Each of the bands was in-gel digested with trypsin and analyzed using Nano-LC/ESI-QTOF-MS. The bands were shown to be acetylated proteins, each with a different number of acetylations and different acetylation sites. The resulting TIC and EIC of the ion at m/z 126.1 of each band are shown in Figure 6. The peptide containing acetylated lysine in each band was clearly located, and the sites of acetylated lysine
Figure 7. Nano-LC/ESI-CID-MS spectra of acetylated peptides a (A), b (B), and c (C) of bands A-C indicated in Figure 6, respectively. KAc, acetylated lysine. 5448 Analytical Chemistry, Vol. 74, No. 21, November 1, 2002
with its amino acid sequence were confirmed by the CID-MS spectrum. (Figure 7). The peaks a-c shown in Figure 6 correspond to the peptides GKAcGGKAcGLGKAcGGAKAcR, GGKAcGLGKAcGGAKAcR, and GLGKAcGGAKAcR of H4, respectively. Band A is identified as tetraacetylated H4 at lysines 5, 8, 12, and 16, band B as triacetylated H4 at lysines 8, 12, and 16, and band C as diacetylated H4 at lysines 12 and 16. The specificity of site occupancy and order of acetylation of these acetylated H4 proteins from HeLa cells treated with TSA are under investigation. CONCLUSION A method of detecting acetylated lysine by analyzing digested protein using HPLC/ESI-MS with a posttranslational modificationspecific marker ion was described, which can identify protein acetylation at specific lysine residues. The ion at m/z 126.1, a fragment ion obtained by CID, was selected as a marker because of its high selectivity and sensitivity. By extracting the marker ion traces in a complex mixture of digested protein, acetylated peptides were located. The acetylation sites and peptide sequences were determined from their CID-MS spectra. The selectivity and sensitivity of the new marker ion at m/z 126.1 was compared with
the acetylated lysine immonium ion at m/z 143.1, which has been used as a marker previously. Because the latter ion overlaps with other internal fragment ions in some peptides, it can potentially cause false-positive results for protein acetylation. In the EIC of the ion at m/z 126.1, no interfering peaks were found. In addition, the new marker ion was ∼9 times more sensitive than the previous one. The novel marker ion was used to identify the acetylation sites of H4 proteins from HeLa cells treated with TSA. This approach will prove useful for identification of specific lysine acetylation in proteins. ACKNOWLEDGMENT This research was supported by a grant (CG1311) from Crop Functional Genomics Center and a grant (PF003101-03) from Plant Diversity Research Center of the 21st Century Frontier Research Program funded by the Ministry of Science and Technology, Republic of Korea. Received for review March 1, 2002. Accepted August 30, 2002. AC0256080
Analytical Chemistry, Vol. 74, No. 21, November 1, 2002
5449
