MS Fragment Ion Coverage of Acidic Residue

Sep 8, 2004 - Sadanori Sekiya,*Yoshinao Wada, andKoichi Tanaka. Koichi Tanaka Mass Spectrometry Research Laboratory, Shimadzu Corporation, 1, ...
0 downloads 0 Views 290KB Size
Anal. Chem. 2004, 76, 5894-5902

Improvement of the MS/MS Fragment Ion Coverage of Acidic Residue-Containing Peptides by Amidation with 15N-Substituted Amine Sadanori Sekiya,*,† Yoshinao Wada,‡ and Koichi Tanaka†

Koichi Tanaka Mass Spectrometry Research Laboratory, Shimadzu Corporation, 1, Nishinokyo-Kuwabaracho, Nakagyo-ku, Kyoto 604-8511, Japan, and Osaka Medical Center and Research Institute for Maternal and Child Health, 840 Murodo-cho. Izumi, Osaka 594-1101, Japan

Tandem mass spectrometry (MS/MS) is a powerful tool for peptide sequencing and characterization. However, the selective cleavage at acidic residues, aspartic acid, and glutamic acid prevents the generation of enough product ions to elucidate the entire sequence. We attempted to solve the problem by converting the residues into the corresponding amides, asparagine and glutamine. The amidation suppressed the cleavage at the converted residues, and the product ions derived from dissociation at other sites became abundant. Incorporation of nitrogen isotope 15N in the amine constituent for amidation minimized the mass change from -0.984 016 to +0.013 019, allowing easy discrimination of acidic and amide residues in the original sequences by MS/MS database search. In addition, the amidated and unchanged peptides had the same nominal mass, even when the transformation was incomplete, which was ∼70% in the current condition. The unmodified acidic residues remaining were rather useful to give marker fragments by the dominant dissociation. These results demonstrate that 15N-amidation is effective in improving the performance of MS/MS to elucidate amino acid sequences of peptides. Mass spectrometry (MS) is a mainstream method for proteomic analysis after striking developments of soft ionization techniques, such as matrix-assisted laser desorption/ionization (MALDI)1-3 and electrospray ionization (ESI).4,5 On the basis of these ionization methods, tandem mass spectrometry (MS/MS) has enabled sensitive and swift elucidation of peptide sequences and, thus, identification of proteins. The MS/MS capability, however, depends on the nature, namely the sequence, of peptides, since it requires the fragmentation of precursor ions. The “mobile proton model”, in which the cleavage of the peptide is initiated by migration of the proton from the initial site * To whom correspondence should be addressed. Phone: +81-75-823-1482. Fax: +81-75-823-3218. E-mail: [email protected]. † Shimadzu Corporation. ‡ Osaka Medical Center and Research Institute for Maternal and Child Health. (1) Tanaka, K.; Waki, H.; Ido, Y.; Akita, S.; Yoshida, Y. Yoshida, T. Rapid Commun. Mass Spectrom. 1988, 2, 151-153. (2) Karas, M.; Hillenkamp, F. Anal. Chem. 1988, 60, 2299-2301. (3) Hillenkamp, F.; Karas, M. Anal. Chem. 1991, 63, 1193A-1203A. (4) Fenn, J. B.; Mann, M.; Meng, C. K.; Won, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (5) Fenn, J. B.; Mann, M.; Meng, C. K.; Won, S. F.; Whitehouse, C. M. Mass Spec. Rev. 1990, 9, 37-70.

5894 Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

of protonation because of the necessity of the proton involvement at a cleavage site for the fragmentation (charge-directed fragmentation), has been proposed for the fragmentation mechanism.6,7 However, if a peptide has a strongly basic group, the charge is sequestered at the basic site, and the cleavage is initiated in the pathway that does not require intramolecular proton transfer (charge-remote fragmentation). Thus, the peptides containing arginine (Arg), for example, require more energy in order to mobilize protons from the basic site for dissociation than does the peptide without basic residues.6 The doubly protonated ions of the basic residue-bearing peptide require less energy for dissociation than the singly protonated ions because one proton is localized at the most basic site, and another proton can work for cleavage. The MS/MS spectra of the peptides with low-energy collisioninduced dissociation (CID) are generally dominated by sequencespecific b- and y-series fragment ions derived from the cleavage at the peptide amide bond.8-13 However, a line of studies has pointed out that the dominant cleavage at particular amino acid residues or residue combination, namely, at the acidic residuess aspartic acid (Asp) and glutamic acid (Glu),14-18 proline,19-21 and histidine,6 for examplesoccurs in certain peptides and, hence, gives poor fragmentation at other amide bonds of their peptide backbones. In these cases, it is difficult for CID to elicit the sequence data amenable to peptide characterization. Furthermore, the presence of Arg with the acidic residues in the peptide is implicated in the dominant cleavage, and the dominant cleavage at Asp and Glu occurs when the number of ionizing protons does not exceed the number of Arg in the peptide. If the peptide lacks Arg or if the number of protons exceeds that of Arg, the available (6) Wysocki, V. H.; Tsaprailis, G.; Smith, L. L.; Breci, L. A. J. Mass Spectrom. 2000, 35, 1399-1406. (7) Dongre, A. R.; Jones, J. L.; Somogyi, A.; Wysocki, V. H. J. Am. Chem. Soc. 1996, 118, 8365-8374. (8) Hunt, D. F.; Yates, J. R., III; Shabanowitz, J.; Winston, S.; Hauer, C. R. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 6233-6237. (9) Poulter, L.; Taylor, L. C. E. Int. J. Mass Spectrom. Ion Processes 1989, 91, 183-197. (10) Alexander, A. J.; Thibault, P.; Boyd, R. K.; Curtis, J. M.; Rinehart, K. L. Int. J. Mass Spectrom. Ion Processes 1990, 98, 107-34. (11) Somogyi, A.; Wysocki, V. H.; Mayer, I. J. Am Soc Mass Spectrom. 1994, 5, 704-717. (12) Papayannopoulos, I. A. Mass Spec Rev. 1995, 14, 49-73. (13) Cox, K. A.; Gaskell, S. J.; Morris, M.; Whiting, A. J. Am. Soc. Mass Spectrom. 1996, 7, 522-531. (14) Qin, J.; Chait, B. T. J. Am. Chem. Soc. 1995, 117, 5411-5412. 10.1021/ac049374r CCC: $27.50

© 2004 American Chemical Society Published on Web 09/08/2004

mobile proton gives the charge-directed fragmentation at various sites of the peptide backbone.6,22,23 To explain the dominant cleavage at the acidic residues in such cases, the side-chain interaction model between Arg and the acidic residue and a cyclic formation between the side chain of the acidic residue and the peptide backbone have been proposed.6,18,22-26 The acidic hydrogen of the carboxyl group of Asp forms a hydrogen bond with the adjacent carbonyl oxygen of the peptide bond (Scheme 1A), the nucleophilic attack of the oxygen of the carboxyl group on the carbon of the peptide forms a five-membered ring (Scheme 1B), and a proton transfers to the amide nitrogen and leads to the cleavage of the amide bonds (Scheme 1C). Wysocki et al. suggested that in the peptides containing both Arg and acidic residues, Arg tightly binds the proton and allows the side-chain of the acidic residue to react easily with the peptide backbone and to form the cyclic peptide that initiates the cleavage.6 Such a dominant cleavage at the acidic residues results in poor production of the fragment ions, except for those derived from the acidic residues and, as a result, prevents the identification of the protein by MS/MS database search. Amidation is a method to modify carboxyl groups by converting the carboxylic acid to amide in the presence of amine constituents,27-29 and the activation of carboxyl group by water-soluble carbodiimides enables mild modification. When ammonia is used as a source of amine, the reaction substitutes glutamine (Gln) and asparagines (Asn) in peptides for Glu and Asp, respectively, and gives a mass decrease of 0.984 016 (monoisotopic mass) per carboxylic acid. When the nitrogen isotope, 15N, is used for amidation, the resulting peptide mass changes by only +0.013 019 and remains within the same nominal mass. This trick would alleviate some confusion due to substitution in the MS/MS database search. In this paper, we show that amidation suppresses the dominant cleavage at the acidic residues and improves the fragment ion coverage. The merits of using 15N in the amine constituent for amidation are also discussed. (15) Tsaprailis, G.; Nair, H.; Somogyi, A.; Wysocki, V. H.; Zhong, W.; Futrell, J. H.; Summerfield, S. G.; Gaskell, S. J. J. Am. Chem. Soc. 1999, 121, 51425154. (16) Huang, Y.; Wysocki, V. H.; Tabb, D. L.; Yates, J. R., III. Int. J. Mass Spectrom. 2002, 219, 233-244. (17) Yu, W.; Vath, J. E.; Huberty, M. C.; Martin, S. A. Anal. Chem. 1993, 65, 3015-3023. (18) Bakhtiar, R.; Wu, Q.; Hofstadler, S. A.; Smith, R. D. Biol. Mass Spectrom. 1994, 23, 707-710. (19) Breci, L. A.; Tabb, D. L.; Yates, J. R., III.; Wysocki, V. H. Anal. Chem. 2003, 75, 1963-1971. (20) Loo, J. A.; Edmonds, C. G.; Smith, R. D. Anal. Chem. 1993, 65, 425-438. (21) Vaisar, T.; Urban, J. J. Mass Spectrom. 1998, 33, 505-524. (22) Tsaprailis, G.; Somogyi, A.; Nikolaev, E. N.; Wysocki, V. H. Int. J. Mass Spectrom. 2000, 195/196, 467-479. (23) Gu, C.; Tsaprailis, G.; Breci, L.; Wysocki, V. H. Anal. Chem. 2000, 72, 58045813. (24) Schlosser, A.; Lehmann, W. D. J. Mass Spectrom. 2000, 35, 1382-1390. (25) Summerfield, S. G.; Whiting, A.; Gaskell, S. J. Int. J. Mass Spectrom. Ion Processes 1997, 162, 149-161. (26) Summerfield, S. G.; Cox, K. A.; Gaskell, S. J. Am. Soc. Mass Spectrom. 1997, 8, 25-31. (27) Hoare, D. G.; Koshland, D. E., Jr. J. Biol. Chem. 1967, 242, 2447-2453. (28) Armstrong, J. M.; Mckenzie, H. A. Biochim. Biophys. Acta 1967, 147, 9399. (29) Rao, M. J.; Acharya, A. S. Methods Enzymol. 1994, 231, 246-267.

Scheme 1. Proposed Mechanism of the Intramolecular Interaction Leading to the Predominant Cleavage at Asp in Peptides

EXPERIMENTAL SECTION Ammonium chloride composed of 14N or 15N, bovine serum albumin (BSA), apomyoglobin from equine skeletal muscle, 1-ethyl-3(3-dimethylaminopropyl)-carbodiimide hydrochloride, R-cyano-4-hydroxycinnamic acid (CHCA), iodoacetamide, bradykinin fragment (1-7), and angiotensin II were purchased from SigmaAldrich (St. Louis, MO). Recrystallized 2,5-dihydroxybenzoic acid (DHB) was purchased from LaserBio Labs (Sophia-Antipolis, France). Guanidine hydrochloride was purchased from ICN Biomedicals Inc. (Irvine, CA). Modified trypsin was purchased from Promega Corp. (Madison, WI). Ammonium bicarbonate, calcium chloride, and acetonitrile were purchased from Wako Pure Chemical Industries Ltd. (Osaka, Japan). Trifluoroacetic acid was purchased from Nacalai Tesque Inc. (Kyoto, Japan). ZipTipµ-C18 and Microcon were purchased from Millipore Co. (Bedford, MA). DL-dithiothreitol was purchased from Fluka Biochemika (Buchs, Switzerland). The synthesis of the apomyoglobin fragment peptide (17-VEADIAGHGQEVLIR-31) was produced by Shimadzu Corp. (Kyoto, Japan). Neurokinin A and substance P were purchased from Peptide Institute, Inc. (Osaka, Japan). Amidation of Proteins. Each 600 pmol of BSA or apomyoglobin dissolved in 50 µL of 5 M guanidine hydrochloride containing 1 M NH4Cl, pH 3.8, was mixed with 15 µL of 5 M guanidine hydrochloride, pH5.5, containing 0.4 M 1-ethyl-3(3dimethylaminopropyl)-carbodiimide hydrochloride (EDAC). After a 2-h incubation at room temperature, the reaction mixture was dialyzed against 0.001 N HCl with Microcon. NH4HCO3 (50 mM final concentration) was then added to the sample solution. The reduction and alkylation were carried out on BSA with DLdithiothreitol (DTT) and iodoacetamide. DTT was added (final concentration was 10 mM) and incubated for 1 h at 56 °C. Iodoacetamide was added (final concentration was 55 mM) and incubated for 45 min at room temperature, then the reduced and alkylated protein was digested with a 1/100 molar ratio of trypsin in the presence of 5 mM CaCl2 for 16 h at 37 °C, and the tryptic digestion was terminated by adding an aliquot of 0.1% trifluoroacetic acid (TFA). The desalting was carried out with ZipTipµC18, and the peptides were eluted in a solution of 0.05% (v/v) TFA and 70% (v/v) acetonitrile (ACN). Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

5895

Figure 1. MS spectra of tryptic digests of amidated and unamidated BSA by AXIMA-CFR plus. The m/z values are given beyond individual signals. Open circles indicate that the ions are actually derived from the BSA peptides. (A) Without amidation. (B) Amidation with 14NH4Cl. (C) Close-up view (upper for unamidated and lower for amidated species) of four representative peaks. Theoretical molecular mass of each peak is indicated at the top. The numbers of D (Asp) and E (Glu) in the peptides are shown in insets.

Amidation of Peptide. The synthetic peptide (50 pmol) was amidated under the same condition described above and desalted with ZipTipµ-C18. Mass Spectrometry. The MS spectra and MS/MS spectra were acquired with AXIMA-CFR plus, MALDI TOF-MS, and AXIMA-QIT, MALDI quadrupole-ion trap TOF-MS instruments (Shimadzu Corp.), respectively. For MS, the peptide solution (1 µL) was mixed with 0.5 µL of a solution of 10 mg/mL CHCA 5896 Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

dissolved in a 0.05% (v/v) TFA and 50% (v/v) ACN solution on a MALDI sample target and then dried. For MS/MS, the peptide solution (1 µL) was mixed with 0.5 µL of 12.5 mg/mL recrystallized DHB dissolved in a 0.05% (v/v) TFA and 50% (v/v) ACN solution on a MALDI sample target and then dried. The monoisotopic mass composed of the lowest mass isotope ion was used. Therefore, “monoisotopic ion” in this paper means the ion composed of the lowest mass isotope.

Figure 2. MS/MS spectra of tryptic peptides of amidated and unamidated BSA by AXIMA-QIT. Letter “I” means internal fragment ions. (A) The product ion mass spectrum from the unamidated precursor at m/z 1479.84. (B) The product ion mass spectrum from the 14N-amidated precursor at m/z 1478.88. The precursor ions for the MS/MS analyses of A and B are depicted in C and D, respectively.

RESULTS AND DISCUSSION Amidation of Protein. Tryptic digest of BSA before and after amidation is shown in Figure 1. Amidation of the carboxyl group with 14N in the amine constituent theoretically causes a mass decrease of 0.984 016. In fact, a decrease of 1 Da at every acidic residue, Asp or Glu, in the peptide was observed, demonstrating that the amidation reaction occurred (Figure 1A-C). The efficiency of reaction was 60-80% by the current condition, as calculated from the isotopic distribution of the peaks; for example, those with molecular mass of 926.49 (65%) and 1438.80 (80%) in Figure 1C. The signal-to-noise ratio (S/N) was improved in many peptides after amidation, for example, a gain of three times for the peak at m/z 1439.92. This was probably due to facile cleavage by trypsin or efficient ionization of the converted peptides. The MS/MS analyses of a tryptic peptide (421-LGEYGFQNALIVR-433, theoretical mass 1478.79) from unamidated BSA and a counterpart peptide derived from amidation with 14NH4Cl were carried out (Figure 2). In the isotopic cluster for the amidated peptide shown in Figure 1C, the ion with m/z 1479.97 was derived from a mixture of amidated and unamidated peptides, whereas

that at m/z 1478.94 was an amidated species. To simplify the evaluation, only the monoisotopic ion was selected as a precursor, as shown in Figure 2C,D. In the MS/MS spectrum of the unamidated peptide, the peak at m/z 1180.69, corresponding to the y10 ion derived from the cleavage at the carboxyl side of Asp, was predominantly detected, and the other fragment ions were much less abundant (Figure 2A). In contrast, for the amidated peptide, the intensity of the y10 ion was not so high, and many other product ions were more detectable (Figure 2B). This result suggested that the suppressed dissociation at this residue increased the generation of other product ions. Apomyoglobin was then analyzed. As expected, the decrease of peptide mass at every 1 Da according to the number of Asp and Glu was observed for most peptides (Figure 3A-C). However, for a peptide with molecular mass of 1814.90 (positions 1-16 of apomyoglobin) which contained two acidic residues, the expected signals after amidation were very small, and instead, prominent signals appeared in the -18 Da region (Figure 3D). The MS/MS Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

5897

Figure 3. MS spectra of tryptic digests of amidated and unamidated apomyoglobin by AXIMA-CFR plus. The m/z values are given beyond individual signals. Open circles indicate that the ions are actually derived from the apomyoglobin peptides. (A) Without amidation. (B) Amidation with 14NH4Cl. (C) Amidation with 15NH4Cl. (D) Close-up view (upper for unamidated and lower for amidated species) of four representative peaks. Theoretical molecular mass of each peak is indicated at the top. The numbers of D (Asp) and E (Glu) in the peptides are shown in insets. 5898 Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

Figure 4. MS/MS spectra of tryptic peptides of amidated and unamidated apomyoglobin by AXIMA-QIT. Letter “I” means internal fragment ions. (A) The product ion mass spectrum from the unamidated precursor at m/z 1606.90. (B) The product ion mass spectrum from the 14Namidated precursor at m/z 1603.95. (C) The product ion mass spectrum from the 15N-amidated precursor at m/z 1606.99. The precursor ions for the MS/MS analyses of A, B, and C are depicted in D, E, and F, respectively.

analysis confirmed that the latter signals were derived from the same peptide (data not shown). A comparison with the mass spectrum of the 15N-incorporated peptide suggested that the peak at m/z 1797.07 of the 14N-incorportaed peptide had only one acidamide conversion and was presumably a dehydrated species of the peptide identified at m/z 1815.09 (Figure 3D). Asp and Glu are known to form a succinimide and a oxazolone by dehydration under certain conditions, and the fully amidated peptides are free from dehydration. This side reaction is discussed below in more detail. The MS/MS analyses of the peptide 17-VEADIAGHGQEVLIR31 (theoretical mass 1605.85) derived from apomyoglobin and the

14N-

and 15N-amidated counterparts were then carried out by selecting monoisotopic ions as precursors (Figure 4). The ion at m/z 1606.99 of the 15N-incorporated sample (Figure 4F) represented a mixture of amidated and unamidated peptides, since the modification was partial, as described above. The ion at m/z 1603.95 of the 14N-incorporated sample (Figure 4E) was a fully amidated species. Comparison of these MS/MS spectra disclosed that the product ions, y11 as well as y4 and y13, in Figure 4C were more abundant than those in Figure 4B. This was due to the fact that y4, y11, and y13 product ion signals in Figure 4C were in part derived from the unamidated peptide, which coexisted with the 15N-amidated peptide and had acidic residues in its sequence. Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

5899

Figure 5. MS spectra of the unamidated and amidated synthetic apomyoglobin peptide corresponding to 17-VEADIAGHGQEVLIR-31 (theoretical molecular mass 1605.85) by AXIMA-CFR plus. (A) Unamidation. (B) Amidation with 14NH4Cl. (C) Amidation with 15NH4Cl. Table 1. The Number of Fragment Ion Series Identified in the MS/MS Spectra of the Tryptic Peptides from Unamidated and 15N-Amidated Apomyoglobin Shown in Figure 4a

a

ion

unamidated

N-amidation

a b c x y z IFb total

3 7 1 0 10 0 14 35

4 7 3 0 11 0 30 55

The mass range is from m/z 400 to m/z 1650. b Internal fragment.

These marks of incomplete amidation in the 15NH4Cl-treated peptides would not be undesirable but, rather, would help to deduce the original sequence. 5900 Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

It was also noted that a variety of ammonium ions were detected in the MS/MS spectra of the amidated peptides. The number of the a-, b-, c-, x-, y-, and z-series and internal fragment ions with significant intensity (S/N > 5) in the MS/MS spectrum was then compared between the unamidated (Figure 4A) and 15Namidated (Figure 4C) peptides. As summarized in Table 1, the amidated peptide generated more various types of fragments than the unamidated one, and the difference was remarkable for the internal fragments. It might be considered that the absence of the acidic residues inducing the facile fragmentation results in promotion of the fragmentation at another peptide backbone and serves to improve the fragment ion coverage in the MS/MS analysis. Amidation of Peptide. When peptides are subjected to the amidation, the C-terminal carboxyl as well as the side carboxyls of Asp and Glu would be converted. To confirm this, a peptide corresponding to 17-VEADIAGHGQEVLIR-31 (theoretical mass,

Figure 6. MS/MS spectra of the synthetic peptides VEADIAGHGQEVLIR from apomyoglobin (theoretical mass 1605.85) by AXIMA-QIT. Letter “I” means internal fragment ions. (A) The product ion mass spectrum from the unamidated precursor at m/z 1607.08. (B) The product ion mass spectrum from the 14N-amidated precursor at m/z 1602.98. (C) The product ion mass spectrum from the 15N-amidated precursor at m/z 1606.90. The precursor ions for the MS/MS analyses of A, B, and C are depicted in D, E, and F, respectively.

1605.85) of apomyoglobin was synthesized and analyzed (Figure 5). Indeed, amidation with 14NH4Cl gave a signal at m/z 1602.97, four units smaller than the MH+ of unmodified peptide, indicating that the C-terminal carboxyl group was amidated (Figure 5B). The MS/MS analyses were performed by selecting the monoisotopic ion as a precursor. The y4, y11, and y13 ions from dissociation at the formerly acidic residues were suppressed by amidation (Figure 6A-C), as in the case of the corresponding tryptic peptide shown in Figure 4. However, when compared with the 14N-amidated tryptic peptide shown in Figure 4B, all y-series fragment ions in

Figure 6B showed a decrease by 1 unit, whereas all b-series fragment ions were unchanged. This was obviously due to amidation of the C-terminal carboxyl group. The number of the a-, b-, c-, x-, y-, and z-series and internal fragment ions with significant intensity (S/N > 5) was increased by amidation, as shown in Table 2. Comparing the MS/MS spectrum of the synthetic apomyoglobin peptide (Figure 6C) with that of the corresponding tryptic peptide (Figure 4C), the product ions, b4, b7, c13, and c14, were detectable only in the former. Easy mobility of protons on the Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

5901

Table 2. The Number of Fragment Ion Series Identified in the MS/MS Spectra (Figure 6) of the Synthetic Peptides before and after 15N-Amidationa

a

ion

before amidation

after N-amidation

a b c x y z 1Fb total

3 9 1 0 11 0 22 46

4 9 5 0 10 0 27 55

The mass range is from m/z 400 to m/z 1650. b Internal fragment.

C-terminally amidated peptide facilitated the detection of these product ions. The MS/MS ion search with Mascot software (Matrix Science, U.K.) normally elicited the apomyoglobin as the first candidate when applying the product ions of 15N-amidated peptide shown in Figure 6C, and ∼20% of the product ions were assigned (data not shown). This showed that the mass shift of +0.013 019 per conversion did not seriously affect the database search. However, the score of the amidated peptide was lower than that of the unamidated peptide, despite the increased number of the matched fragment ions. Because Mascot software performs a data processing of the entered mass data, it is considered that the increase of internal fragment ions and the change of the peak intensity in the MS/MS spectrum of 15N-amidated peptide, which contained substitution of amide residues for acidic ones, are not suited for the data processing algorithm. A database software that requires no consideration of peak intensity and the number of internal fragment would allow maximizing the effect of 15N-amidation. Nature of the Decrease of 18 Da by Amidation. In some cases, amidation generated the peptides with a mass of 18 Da less than the expected products. As described above, this (30) Geiger, T.; Clarke, S. J. Biol. Chem. 1987, 262, 785-794. (31) Athmer, L.; Kindrachuk, J.; Georges, F.; Napper, S. J. Biol. Chem. 2002, 277, 30502-30507. (32) Reissner, K. J.; Aswad, D. W. Cell. Mol. Life Sci. 2003, 60, 1281-1295. (33) Sewald, N.; Jakubke, H. D. Peptides: Chemistry and Biology; Wiley-VCH: Weinheim, Germany, 2002; pp 191-195.

5902 Analytical Chemistry, Vol. 76, No. 19, October 1, 2004

phenomenon was most prominent for the tryptic apomyoglobin peptide of 1814.90 Da (Figure 3D) and, to a lesser extent, for the synthetic peptide in Figure 5B and C. To address the nature of such a mass shift, a bradykinin fragment (RPPGFSP), neurokinin A (HKTDSFVGLM-NH2) and substance P (RPKPQQFFGLM-NH2) were subjected to the amidation reaction. The species with mass decrease was marked in the bradykinin fragment (data not shown), and appeared at the addition of EDAC to the sample. The peak was less abundant for neurokinin A and was undetectable for substance P. These results suggested that the activation of carboxyl groups, largely at the C-terminal, by carbodiimide were involved in the formation. The decrease of 18 Da was most probably due to dehydration that occurred between the carboxyl and amino groups. In addition, the activation of the carboxyl group by carbodiimide might lead to a cyclization between the carboxyl group and the peptide bond, forming succinimide or oxazolone. It is known that succinimide is formed through deamidation of asparagine or dehydration of aspartic acid by the nucleophilic attack of the peptide bond nitrogen on the side chain carbonyl carbon,30-32 and the oxazolone formation occurs by nucleophilic attack of the acyl group on the side chain carbonyl carbon when the carboxyl of the side chain is activated.33 Moreover, it is suggested that amidation has the possibility to form links between the carboxyl group of a molecule and the amino group of another molecule, leading to polymers (intermolecular interaction). The protection of the amino groups in the protein or peptide would be effective to resolve this unfavorable reaction. Indeed, acetylation of the amino groups of angiotensin II successfully suppressed dehydration (data not shown). CONCLUSION Amidation is a useful modification to improve the fragment ion coverage in the MS/MS analysis of the peptide containing acidic residues. The mass shift caused by this modification is compensated with the use of 15N in the amine constituent. Modification of the data-processing algorithm for the MS/MS database search will unravel the true performance of this modification. Received for review April 27, 2004. Accepted August 3, 2004. AC049374R