In-Gel Derivatization of Proteins for Cysteine-Specific Cleavages and their Analysis by Mass Spectrometry Mario Thevis,*,† Rachel R. Ogorzalek Loo,‡ and Joseph A. Loo*†,‡,§ Departments of Chemistry and Biochemistry, Biological Chemistry, and the Mass Spectrometry and Proteomics Technology Center, University of California, Los Angeles, California 90095-1570 Received September 16, 2002
As a potential tool for proteomics and protein characterization, in-gel cysteine- and arginine-specific cleavage is demonstrated by means of trypsin or endoproteinase Lys-C for six model proteins (lysozyme, R-lactalbumin, β-lactoglobulin, ribonuclease A, albumin, and transferrin), ranging in size from 14 kDa to 79 kDa. Chemical modifications of cysteine (aminoethylation with bromoethylamine or N-(iodoethyl)trifluoroacetamide, and subsequent guanidination) and lysine (acetylation) prior to tryptic digestion releases peptides delineated by cysteine or arginine residues. Peptide products are analyzed by MALDITOF-MS, ESI-MS, and ESI- and MALDI-MS/MS (with a quadrupole time-of-flight instrument). Complications induced by acrylamide alkylations of cysteines were avoided by substituting lower pH bis-tris polyacrylamide gels for tris-glycine. Sequence coverages from 35 to 86% were obtained and amino acid compositions of generated peptides could be confirmed by comprehensive y- and b-ion series. Detailed information about, in particular, cysteine rich proteins after gel electrophoresis were obtained. The chemistries for modification and cleavage specificities at cysteine residues provide an alternative means to characterize and identify proteins separated by gel electrophoresis. Keywords: cysteine cleavage • MALDI-MS/MS • ESI-MS • bromoethylamine • QTOF • lysine
Introduction Enzymatic degradation of proteins at basic amino acid residues such as arginine and lysine by trypsin is a common technique for protein characterization and identification. Numerous applications of in-solution and in-gel digests have been reported, providing detailed information on proteins, especially when combined with MALDI-MS or ESI-MS with tandem mass spectrometry (MS/MS).1 The quest for procedures or reactions enabling amino acid-specific cleavages in proteins maintains high popularity, especially because the scientific potential of proteomics has been realized. Thus, several chemical reactions have been established and several enzymes have been discovered that degrade proteins site-specifically within the sequence; comparison of the experimentally observed peptide fragments with those predicted from sequence databases establishes the means to identify proteins. For instance, proteins are cleaved at aspartic acid residues by incubation at 108 °C in 2% formic acid,2 after methionines utilizing cyanogen bromide3 or after tryptophan residues by means of BNPS* To whom correspondence should be addressed. Mario Thevis and Joseph A. Loo, University of California at Los Angeles, Department of Chemistry and Biochemistry, 416 Paul D. Boyer Hall (MBI), 611 Charles E. Young Drive East, Box 951570, Los Angeles, CA, 90095-1570. Tel. (310) 7947308. Fax (310) 206-7286. Email:
[email protected] (M. Thevis).
[email protected] (J. A. Loo). † Department of Chemistry and Biochemistry, University of California. ‡ Department of Biological Chemistry, University of California. § Department of Mass Spectrometry and Proteomics Technology Center, University of California. 10.1021/pr025568g CCC: $25.00
2003 American Chemical Society
skatole4,5 or o-iodobenzoic acid.6 Enzymes with different specificities such as pepsin, chymotrypsin, Glu-C, and a variety of others are employed routinely for protein characterization. Already in 1956, Lindley7 described the possibility to derivatize reduced cysteine residues in proteins with 2-bromoethylamine (BEA), to imitate lysine’s structure and thus make cysteines accessible to trypsin. Because aminoethylated cysteine and lysine differ only in a methylene group replaced by a sulfur atom (Figure 1), trypsin proved to be capable of cleaving modified cysteines at their C-terminus. In the following years, solution-based procedures were developed according to this observation,8,9 and alternative methods to derivatize cysteine were demonstrated.10,11 The separation of complex protein mixtures is commonly done by means of one- or two-dimensional gel electrophoresis, and in conjunction with mass spectrometry-based identification procedures, it forms the basis of a commonly employed strategy for proteomics. Therefore, additional chemistries and cleavage specificities for the analysis of gel-embedded proteins may offer unique advantages and tools to proteomics-based research. The potential of cysteine reactive cleavages of proteins of various sizes separated by gel electrophoresis was evaluated. A key challenge in exploiting cysteine chemistry was to circumvent the residue’s reactivity with free acrylamide, achieved here by electrophoresing on lower pH-formulated polyacrylamide gels. Combined with acetylation of lysine residues, we are able to achieve selective cleavages at modified cysteines and arginines utilizing trypsin, and solely at cysteines utilizing Journal of Proteome Research 2003, 2, 163-172
163
Published on Web 12/24/2002
research articles
Figure 1. Chemical structures of S-aminoethyl cysteine (A) and lysine (B).
endoproteinase Lys-C. Although arginine-specific degradation of proteins is possible with the enzymes Arg-C and Clostripain, mainly in-solution applications have been established. Because chemically blocked lysine residues are not cleaved by trypsin, an alternative to Arg-C for digesting proteins in-gel C-terminal to arginine is also demonstrated.
Experimental Section Chemicals and Proteins. Acetic anhydride (98%), ammonium bicarbonate (99%), 2-bromo ethylamine hydrobromide, lysozyme, R-lactalbumin, β-lactoglobulin, ribonuclease A, and human apo-transferrin were purchased from SIGMA (St. Louis, MO). Recombinant human serum albumin was obtained from NEW CENTURY PHARMACEUTICALS (Huntsville, AL). Methanol and acetonitrile (both HPLC grade) were purchased from MERCK (Darmstadt, Germany). O-methyl isourea hemisulfate (94%) was from ACROS (Fair Lawn, NJ) and dithiothreitol from BIO-RAD (Hercules, CA). The enzymes modified trypsin (sequence grade) and lysyl endopeptidase were purchased from PROMEGA (Madison, WI) and WAKO (Osaka, Japan), respectively. All buffers and solutions were prepared using deionized water (MilliQ grade). Gel Electrophoresis. Approximately 20 to 250 pmol of the proteins were loaded onto tris-glycine (4-12%) or NuPage bistris (10%) (both INVITROGEN, Carlsbad, CA) pre-cast polyacrylamide gels (1 mm thickness), and one-dimensional gel electrophoresis was performed under reducing conditions using tris-glycine SDS running buffer or 3-(N-morpholino) propane sulfonic acid (MOPS) as running buffer, respectively. The bistris gels employed the NuPage formulations of LDS sample buffer (lithiumdodecyl sulfate), reducing agent (DTT), and running buffer anti-oxidant, whereas tris-glycine gels used Laemmli sample buffer (including sodium dodecyl sulfate and 2-mercaptoethanol) purchased from SIGMA. After fixing, staining was done with GelCode Blue (PIERCE, Rockford, IL) stain reagent according to the recommended protocol. Mass Spectrometry. Mass spectra of chemically modified and digested proteins were recorded on an Applied Biosystems (Foster City, CA) Voyager DE STR MALDI-TOF mass spectrometer and a QStar Pulsar I (QqTOF) equipped with a MALDI source or a nanoelectrospray (PROTANA, Odense, Denmark) interface. With MALDI, R-cyano-4-hydroxy cinnamic acid (saturated solution) or 2,5-dihydroxybenzoic acid (20 mg/mL) in 50% acetonitrile containing 0.1% TFA were used as matrix. Product ion spectra for sequence determination of peptides were generated by the QqTOF mass spectrometer using nitrogen or argon as collision gases with ESI or MALDI, respectively. In-Gel Derivatization and Digestion of Proteins. In-Gel Acetylation. Coomassie-stained gel bands were excised and transferred to microcentrifuge tubes. Destaining was accomplished by incubation in 100 µL of a mixture of 100 mM 164
Journal of Proteome Research • Vol. 2, No. 2, 2003
Thevis et al.
ammonium bicarbonate (aqueous) and acetonitrile (1:1, v:v) at room temperature for 15 min. The supernatant was discarded, and the gel piece was dehydrated in 30 µL of neat acetonitrile. After 10 min, the supernatant was discarded and the gel was dried in a vacuum centrifuge. The dry gel was swollen in 20 µL of 50 mM ammonium bicarbonate, and 50 µL of 30% acetic anhydride in methanol was added to acetylate reactive lysine side-chains (and the N-terminus). The samples were kept at room temperature for 2 h, the supernatant was removed and the gels were washed twice with 70 µL of 100 mM ammonium bicarbonate. Finally, the gels were dehydrated in 30 µL of neat acetonitrile and dried in a vacuum centrifuge. For arginine-specific degradation, the gels were then treated following commonly accepted protocols for tryptic digestion (e.g. Shevchenko et al.12), whereas cysteine and arginine or solely cysteine cleavages employed the following additional preparative steps. In-Gel Aminoethylation of Cysteine. The gel entrapped, acetylated proteins were reduced by reswelling the gel in 30 µL of 10 mM dithiothreitol (DTT) in 500 mM sodium carbonate at 60 °C. After 1 h, 10 mg of 2-bromo ethylamine hydrobromide (BEA) dissolved in 10 µL of 500 mM sodium bicarbonate was put into the tube, and the pH was adjusted to 10 by means of 100 mM sodium hydroxide. Maintaining the sample at 50 °C, control and adjustment of the pH to 10 was repeated three times within the first hour and the reaction was continued for another 3 h. Afterward, the supernatant was discarded, the gel was washed twice with 50 mM ammonium bicarbonate, dehydrated in acetonitrile and dried in a vacuum centrifuge. Subsequently, the digest of the proteins was performed by the addition of 60 µl of 50 mM ammonium bicarbonate containing 1 µg of trypsin or endoproteinase Lys-C, and incubating the sample at 37 °C overnight. The resulting peptides were extracted from the gel twice with 50 µL of 50% acetonitrile containing 1% TFA. The combined extracts were dried in a vacuum centrifuge, and either reconstituted in 0.1% TFA and desalted by C-18 ZipTip (Millipore, Bedford, MA) pipetting, or subjected to further derivatization. Additional modification of peptides to their guanidinated counterparts was achieved by solving the extracted and dried peptides in 30 µL of 2M O-methylisourea hemisulfate in 250 mM sodium carbonate and heating at 40 °C for 90 min.13-15 Subsequently, the derivatized peptides were withdrawn and prepared for analysis by utilizing C-18 ZipTips.
Results and Discussion In-Gel Derivatization of Proteins. Acetylating lysine and aminoethylating cysteine prior to trypsin digestion generates peptide fragments originating from specific cleavage C-terminal to cysteine and arginine. In the present study, we investigated the ability to derivatize and digest six different proteins (chicken egg lysozyme, bovine R-lactalbumin, bovine β-lactoglobulin, bovine ribonuclease A, human serum albumin, and human apo-transferrin) after 1-D gel electrophoresis. The acetylation of lysine side-chains ensures the resistance of this residue to tryptic cleavage and, thus, limits the number of small peptides generated by the additional cysteine cleavage sites. The conditions described quantitatively modify the primary amino group of lysines and, in addition, the Nterminus of the protein. Hence, no peptides originating from a C-terminal cleavage after lysine were observed in any of the investigated proteins, and the acetylation of lysines was found to be complete in all peptides examined by MS/MS.
research articles
In-Gel Derivatization of Proteins
Figure 2. MALDI-TOF mass spectrum of chicken egg lysozyme after acetylation of lysine side-chains, aminoethylation of cysteines and trypsin digestion. Table 1. Peptides Calculated for Chicken-Egg Lysozyme after Cysteine and Lysine Modification and Subsequent Trypsin Digestion peptide no.
AA residues
sequence
MH+cal
MH+exp
1 2 3 4 5a 6 7 8 9
15-21 7-14 22-30 116-125 115-125 81-94 31-45 46-61 95-112
HGLDNYR ELAAAMKR GYSLGNWVC KGTDVQAWIR CKGTDVQAWIR SALLSSDITASVNC AAKFESNFNTQATNR NTDGSTDYGILQINSR AKKIVSDGNGMNAWVAWR
874.93 932.07 1042.14 1216.39 1362.55 1423.55 1741.87 1754.85 2088.41
874.85 n.d. 1042.06 1216.29 1362.00 1423.76 1741.78 1754.78 2088.30
a
Missed cleavage at Cys 115.
Aminoethylation of sulfhydryl groups with BEA required, that the pH be adjusted during the first hour of the reaction, because the BEA hydrobromide salt releases an equivalent of HBr for every modified cysteine. Excess of BEA quantitatively converts cysteines to their S-aminoethyl derivatives, but its presence for prolonged reaction times at elevated temperatures affects primary amino groups, as well. Because the primary amino groups of lysines are already acetylated in this protocol, the only primary amines amenable to modification are those
attached to the cysteine residues, and such a side reaction would hamper subsequent tryptic cleavage. In this regard, other undesired reactions such as acrylamide adduct formation at cysteine residues16-19 were to be avoided, as they, too, would inhibit the aminoethylation of cysteines and, thus, the intended cleavage at these particular amino acids. Hence, the use of NuPage bis-tris polyacrylamide gels with MOPS as a running buffer was preferred over tris-glycine gels, for instance, which showed substancial capability to produce acrylamide adducts with proteins. As a consequence, peptides bearing missed cleavages at non-aminoethylated cysteines were observed for proteins separated with tris-glycine gels (data not shown) but not with bis-tris gels. The absence of cysteine S-beta propionamide (acrylamide adduct) in bis-tris gels is attributed to their lower pH environment.20 Tris-glycine Laemmli-type gels are cast at pH 8.7 and achieve a pH of approximately 9.5 in the resolving region during electrophoresis, whereas bis-tris gels, cast at pH 6.4, operate at pH 7.0. An alternative reagent for the aminoethylation of proteins and peptides is N-(iodoethyl)trifluoroacetamide9,21 (e.g., Aminoethyl-8, PIERCE). Its advantage over BEA is that it protects the amino group by a trifluoroacetyl during the sulfhydryl derivatization. Thus, additional ethylamine residues will not attach to modified cysteines. This trifluoroacetyl group is removed under alkaline conditions, yielding the desired aminoethylated cysteines. However, using this reagent, we observed fewer peptides in comparison to samples prepared with BEA, and those peptides common to mass spectra of both samples were observed with a higher abundance with the use of BEA (vide infra). Mass Spectrometry of Protein Cleavage Products. In Figure 2, the MALDI-TOF mass spectrum of chicken egg lysozyme (14306 Da, 8 cysteines) is shown, recorded after acetylation, aminoethylation and trypsin digestion. In Table 1, the peptides generated theoretically by the method described above are listed. The modification of each lysine residue increases the peptide mass by 42 Da, whereas the aminoethylation of cysteine residues elevates the mass by 43 Da. Only peptides greater than 800 Da were analyzed, given the difficulty to detect smaller peptides by MALDI-TOF-MS. The comparison of the mass spectrum of Figure 2 with the predicted peptide masses calculated for cysteine- and lysine-modified lysozyme demon-
Table 2. Peptides Observed after Modification and Subsequent Trypsin Digestion of R-lactalbumin, β-lactoglobulin, and Ribonuclease A
a
No.
AA residues
1 2 3 4 5 6 7
112-120 62-73 78-91 92-108 11-28 92-111 29-61
1 2 3 4 5
107-119 149-162 125-148 125-148 41-66
1 2 3 4
1-10 111-124 96-110 41-58
sequence
lactalbumin SEKLDQWLC KDDQNPHSSNIC DKFLDDDLTDDIMC VKKILDKVGINYWLAH ELKDLKGYGGVSLPEWVC VKKILDKVGINYWLAHKALC TTFHTSGYDTQAIVQNNDSTEYGLFQINNKIWC lactoglobulin MENSAEPEQSLAC LSFNPTQLEEQCHI TPEVDDEALEKFDKALKALPMHIR TPEVDDEALEKFDKALKALPMaHIR VYVEELKPTPEGDLEILLQKWENGEC ribonuclease A KETAAAKFER EGNPYVPVHFDASV AYKTTQANKHIIVAC KPVNTFVHESLADVQAVC
MH+cal
MH+exp
1207.33 1443.47 1744.87 2024.38 2121.38 2524.96 3896.18
1207.54 1443.70 1744.72 2024.49 2121.42 2524.72 3896.04
1452.56 1702.86 2894.27 2910.27 3160.49
1452.15 1703.33 2894.03 2910.01 3160.12
1277.30 1531.66 1788.89 2043.33
1277.12 1531.20 1788.44 2043.65
Oxidized.
Journal of Proteome Research • Vol. 2, No. 2, 2003 165
research articles
Thevis et al.
Figure 3. ESI-QqTOF product ion scans of (A) peptide 22-30 and (B) peptide 31-45 of chicken egg lysozyme after acetylation of lysine side-chains, aminoethylation of cysteines and trypsin digestion.
strates the ability to derivatize the gel-embedded protein in the desired way. Except for the expected signals at m/z 932 and 1424, all peptides theoretically calculated are observed by MALDI-MS, and by means of ESI-MS, also the peptide 8194 (m/z 1424) could be identified and characterized, giving a total sequence coverage of 69%. The peptides containing a 166
Journal of Proteome Research • Vol. 2, No. 2, 2003
C-terminal arginine are more abundant than those having a C-terminal cysteine but besides the fragments 22-30 (m/z 1042) and 81-94 (m/z 1424) which include a modified cysteine, the peptides 116-125 (m/z 1216), 31-45 (m/z 1742), and 95112 (m/z 2088) originate from cleavage at a cysteine residue N-terminal to the peptide. In addition, we observed an intense
In-Gel Derivatization of Proteins
research articles
Figure 4. ESI-QqTOF product ion scan of peptide 92-107 of R-lactalbumin. The entire amino acid sequence between y1 and y14 of the peptide originating from an unspecific cleavage after histidine 107 could be confirmed.
Figure 5. ESI-QqTOF product ion scan of peptide 1-10 of ribonuclease A. The lysine at position one contains two acetyl groups as verified by the abundant b1-ion at m/z 213.
peptide at m/z 1362.6 resulting from a missed cleavage at the modified cysteine 115. The sequence of this peptide was characterized by ESI-MS/MS showing that it contained the
amino acids 115-125 and terminated with arginine. All predicted sequences of the observed peptides could be verified by ESI-MS/MS of the (M+2H)2+ as shown exemplarily for the Journal of Proteome Research • Vol. 2, No. 2, 2003 167
research articles
Thevis et al. Table 3. Peptides Observed after Modification and Subsequent Trypsin Digestion of Human Serum Albumin and Human Apo-Transferrin AA no. residues
Figure 6. MALDI-TOF mass spectra of human serum albumin after acetylation of lysine side-chains and aminoethylation of cysteines by means of (A) 2-bromoethylamine with subsequent trypsin digestion; or (B) Aminoethyl-8 with subsequent trypsin digestion. Comparable peptides are observed in both spectra but at decreased intensities in case of the utilization of Aminoethyl8.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15a 16a
210-218 201-209 338-348 337-348 449-461 102-114 146-160 411-428 393-410 317-336 125-144 223-245 11-24 370-392 82-90 349-360
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
369-377 162-171 476-484 624-632 485-495 242-254 125-137 603-615 144-158 180-194 507-522 309-324 457-474 645-663 378-402 195-220 a
Figure 7. MALDI-TOF mass spectrum of human apo-transferrin after acetylation of lysine side-chains, aminoethylation of cysteines and trypsin digestion. All tagged signals are assigned to peptides of the modified and digested protein.
peptides 22-30 and 31-45 in Figure 3A-B. The C-terminal aminoethylated cysteine is found at m/z 165.05 as the y1-ion (Figure 3A) and the modified lysine residue is calculated with 170 Da (Figure 3B), present in the intense b3-ion at m/z 313.16. Further examples of small proteins (
