Negative Ion Postsource Decay Time-of-Flight Mass Spectrometry of

ACS eBooks; C&EN Global Enterprise. A; Accounts ..... Time-of-flight mass spectrometry (TOF) with matrix-assisted laser ... The effective flight path ...
0 downloads 0 Views 79KB Size
Anal. Chem. 1998, 70, 5122-5128

Articles

Negative Ion Postsource Decay Time-of-Flight Mass Spectrometry of Peptides Containing Acidic Amino Acid Residues Jaran Jai-nhuknan and Carolyn J. Cassady*

Department of Chemistry and Biochemistry, Miami University, Oxford, Ohio 45056

Acidic peptides have been studied by negative ion postsource decay (PSD) matrix-assisted laser desorption/ ionization time-of-flight (MALDI-TOF) mass spectrometry. The peptides contained from 5 to 16 residues and were chosen on the basis of their patterns of the acidic residues. Using typical MALDI sample preparation techniques employing an acidic matrix, gastrin I (1-14), and epidermal mitosis inhibiting pentapeptide yielded much larger deprotonated ion signals, [M - H]-, than protonated ions, [M + H]+. This may be due to their absence of basic residues, coupled with their arrays of acidic residues. The PSD fragmentation of the peptide negative ions showed that an array of acidic residues, as in gastrin I (1-14), yielded simple spectra containing mainly backbone cleavage ions from the C-terminus. Hirudin (5465), which contains two sets of two consecutive Glu residues, and fibrinopeptide A and fibrinopeptide B, with isolated acidic residues, also showed backbone cleavages as common fragment ions. In addition, the two sets of isolated consecutive amino acid residues in Cys(Bzl)84CD4 (81-92) and hirudin (54-56) yielded internal ions from the cleavages at the (OdC)sNH bond between the acidic residues. Also observed were ions with unique side chain losses, such as the loss of C6H4O from a tyrosine residue and SCH2C6H5 and CH2C6H5 from a benzylated cysteine residue. Compared to the positive mode, the negative-ion PSD yielded fewer fragments which usually involved only one type of backbone cleavage (e.g., [yn H2O]-). These simple spectra aided interpretation. Overall, the acidic peptides studied yielded negative ion PSD spectra that were useful for peptide sequencing. Mass spectrometry has become a method of choice in determining the amino acid sequences of peptides. Fragmentation of positive ions has been the subject of many studies.1-4 However, there are classes of peptides that may not yield desirable spectra * Correspond author: (phone) (513)-529-2494; (fax) (513)-529-1667; (e-mail) [email protected]. (1) Papayannopoulos, I. Mass Spectrom. Rev. 1995, 14, 49-73. (2) Biemann, K.; Martin, S. A. Mass Spectrom. Rev. 1987, 6, 1-76. (3) Kaufmann, R. J. Biotechnol. 1995, 41, 155-175. (4) Spengler, B. J. Mass Spectrom. 1997, 32, 1019-1036.

5122 Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

under positive-ion mass spectrometric conditions. Examples are species containing numerous acidic residues such as glutamic acid (Glu) and aspartic acid (Asp), as well as those with modified side chains such as sulfate (OSO3H) and phosphate (OPO3H2). Due to their high acidity, these peptides may form deprotonated pseudomolecular ions, [M - H]-, more readily than protonated ions, [M + H]+. Analysis of natural and posttranslational peptides such as those containing arrays of acidic groups is important. Peptides and proteins in neurological sciences often have a large number of acidic residues.5-8 Several tandem peptides with more than 50% acidic residues are being studied as potential vaccines for malaria.9 Sulfonated or phosphorylated serine (Ser), threonine (Thr), and tyrosine (Tyr) are often found in peptides and proteins upon posttranslation, with phosphorylation being common for enzyme activation or deactivation in nature.10 Therefore, analysis of phosphorylated proteins and polypeptides is essential. Low intensities for [M + H]+ or fragments may plague positivemode studies of acidic peptides. For example, Vinh et al.11 have reported these problems while studying the posttranslational polyglycylation of tubulin. In work by Lapko et al.,12 unsuccessful attempts to analyze phosphorylated peptides in the positive mode were overcome by modifying the phosphate side chain. For sulfated peptides, Yagami et al.13 reported that liquid secondary ion mass spectrometry (LSIMS) of Tyr-O-sulfate nonapeptides yielded intense [M - H]- but significantly less [M + H]+. (5) Pierotti, A. R.; Prat, A.; Chesneau, V.; Gaudoux, F.; Leseney, A.-M.; Foulon, T.; Cohen, P. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 6078-6082. (6) Foulon, T.; Cadel, S.; Chesneau, V.; Draoui, M.; Prat, A.; Cohen, P. In Neuropeptides: Basic and Clinical Advances; Crawley, J. N., McLean, S., Eds.; The New York Academy of Sciences: New York, 1996; pp 106-120. (7) Lemaire, S.; Yamashiro, D.; Rao, A. J.; Li, C. H. J. Med. Chem. 1977, 20, 155-158. (8) Tatemoto, K. Proc. Natl. Acad. Sci. U.S.A. 1982, 79, 5485-5489. (9) Chougnet, C.; Troye-Blomberg, M.; Deloron, P.; Kabilan, L.; Lepers, J. P.; Savel, J.; Perlmann, P. J. Immunol. 1991, 147, 2295-2301. (10) Voet, D.; Voet, J. G. Biochemistry, 2nd ed.; John Wiley & Son, Inc.: New York, 1995. (11) Vinh, J.; Loyaux, D.; Redeker, V.; Rossier, J. Anal. Chem. 1997, 69, 39793985. (12) Lapko, V. N.; Jaing, X.-Y.; Smith, D. L.; Song, P.-S. Biochemistry 1997, 36, 10595-10599. (13) Yagami, T.; Kitagawa, K.; Futaki, S. Rapid Commun. Mass Spectrom. 1995, 9, 1335-1341. 10.1021/ac980577n CCC: $15.00

© 1998 American Chemical Society Published on Web 11/04/1998

Table 1. Peptides Used in This Study

a

peptide

sequencea

source

gastrin I (1-14) epidermal mitosis inhibiting pentapeptide synthetic enteropeptidase substrate hirudin (54-65) fibrinopeptide A fibrinopeptide B Cys(Bzl)84-CD4 (81-92)

H-pEGPWIEEEEEAYGW-OH H-pEEDSG-OH H-GDDDDK-β-NPb H-GDFEEIPEEYLQ-OH H-ADSGEGDFLAEGGGVR-OH H-pEGVNDNEEGFFSAR-OH H-TYIC(Bzl)EVEDQKEE-OH

human mouse synthetic leech Hirudo medicinalis human human human, derivatized

Acidic residues are shown in italic type. b β-NP ) β-NH-C10H7.

Although reports on protonated peptides far outnumber those of deprotonated peptides,1,2 a number of recent studies have focused on the negative-ion spectra of peptides and on MS/MS fragmentation mechanisms.14-20 Bowie14-17 and Beynon20 and their co-workers have reported high-energy collision-induced dissociation (CID) mechanisms of small peptide anions produced by fast atom bombardment (FAB). Fragmentation of di- and tri15,20 and tetrapeptides14 yields backbone cleavages that allow determination of amino acid sequences. Moreover, in larger peptides, the negative-ion spectra are as informative as positiveion spectra.16 Beauchamp and co-workers18 have studied the fragmentation of small negative peptide ions using a FAB/Fourier transform mass spectrometer (FTMS). It was found that, under low-energy CID conditions, fragmentation is initiated by the sites of deprotonation on the acidic side chain or the C-terminal carboxyl group when no acidic residues are present. Yang and Wilkins19 have shown that the sequence information can be obtained from larger peptides (3-15 residues, Mr ) 246-1880) by CID in an FTMS coupled with laser desorption/ionization. Time-of-flight mass spectrometry (TOF) with matrix-assisted laser desorption/ionization (MALDI) has become an important technique in the analysis of biomolecules. For peptides and proteins, MALDI-TOF provides excellent sensitivity. In addition, the 1994 introduction of postsource decay (PSD) metastable fragment analysis on reflectron instruments21 has made MALDITOF attractive for peptide sequencing. Recently, we reported the first negative-ion PSD study of small peptides.22 It was found that negative and positive PSD spectra of small peptides yielded complementary sequence information. Ions incorporating the N-terminus (an+ and bn+) are often seen in the positive mode, while C-terminal fragments (yn- and [yn - 18]-) are common in the negative mode. In this study, we demonstrate the use of negative-ion PSD for analyzing peptides that contain various distributions of acidic (14) Bradford, A. M.; Waugh, R. J.; Bowie, J. H. Rapid Commun. Mass Spectrom. 1995, 9, 677-685. (15) Waugh, R. J.; Bowie, J. H. Rapid Commun. Mass Spectrom. 1994, 8, 169173. (16) Steinborner, S. T.; Bowie, J. H. Rapid Commun. Mass Spectrom. 1996, 10, 1243-1247. (17) Steinborner, S. T.; Bowie, J. H. Rapid Commun. Mass Spectrom. 1997, 11, 253-258. (18) Marzluff, E. M.; Campbell, S.; Rodgers, T. M.; Beauchamp, J. L. J. Am. Chem. Soc. 1994, 116, 7787-7796. (19) Yang, L.-C.; Wilkins, C. L. Org. Mass Spectrom. 1989, 24, 409-414. (20) Bradley, C. V.; Howe, I.; Beynon, J. H. Biomed. Mass Spectrom. 1981, 8, 85-89. (21) Kaufmann, R.; Kirsch, D.; Spengler, B. Int. J. Mass Spectrom. Ion Processes 1994, 131, 355-385.

residues. The chosen peptides, which are listed in Table 1, have arrays of Glu or Asp, two consecutive acidic residues (Glu-Glu or Glu-Asp), or an acidic residue among neutral or basic residues. The results demonstrate that MALDI yields intense negative signals for acidic peptides and that PSD on these anions provides primary structure information. EXPERIMENTAL SECTION The instrument configuration has been discussed previously.22-24 Briefly, all experiments were performed on a Reflex III TOF MS (Bruker Daltonics, Billerica, MA) equipped with a Nd:YAG laser model MiniLase-10 (New Wave Research, Sunnyvale, CA) for MALDI. Initial experiments involved continuous ion extraction (CE) with a single-stage gridless ion source and an accelerating voltage of 30 kV. Later experiments utilized the delayed-extraction (DE) mode with an extraction delay of 350 ns and a 20-kV accelerating voltage. Following extraction from the source, ions traveled through a field-free region into a two-stage gridless reflectron and were detected with a dual-channel plate electron multiplier at the end of a second field-free region. The effective flight path was 290 cm. For PSD experiments, the mass spectrum at each voltage was the sum of 40 laser shots. The laser wavelength was 355 nm. The laser power was maintained near the threshold required to produce ions and was increased if product ion intensity noticeably diminished. Prior to acquiring a PSD spectrum, the instrument was mass calibrated at the maximum reflectron voltage using [M + H]+ or [M - H]- from renin substrate tetradecapeptide (horse) (Mr 1560) and R-cyano-4-hydroxycinnamic acid (CCA). Fibrinopeptide B was used as the negative-ion PSD calibrant23 while ACTH (18-39) was employed to calibrate positive-ion PSD.25 Since this work is the first report on PSD of acidic deprotonated peptides, accurate peak assignment is important. Therefore, to obtain high signal intenstity and mass accuracy, peptide solutions were prepared at a relatively high concentration of 0.1 mg/mL (60-200 µM) in acetonitrile/water (70:30 v/v) with 0.1% trifluoroacetic acid (TFA). The matrix was saturated CCA in the same solvent system. A 1:4 (v/v) ratio of peptide solution to matrix solution was employed. Peptides were purchased from Sigma (St. (22) Jai-nhuknan, J.; Cassady, C. J. Rapid Commun. Mass Spectrom. 1996, 10, 1678-1682. (23) Jai-nhuknan, J.; Cassady, C. J. J. Am. Soc. Mass Spectrom. 1998, 9, 540544. (24) Zhang, X.; Jai-nhuknan, J.; Cassady, C. J. Int. J. Mass Spectrom. Ion Processes 1997, 171, 135-145. (25) Rouse, J. C.; Yu, W.; Martin, S. A. J. Am. Soc. Mass Spectrom. 1995, 6, 822-835.

Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

5123

Figure 1. Comparison of MALDI-TOF spectra in the (a) negativeand (b) positive-ion modes for gastrin I (1-14).

Figure 2. Negative-ion PSD spectrum of gastrin I (1-14).

Louis, MO), Bachem Bioscience Inc. (Philadephia, PA), or ICN Biomedical Research Products (Costa Mesa, CA). Acetonitrile, ammonia, CCA, and TFA were obtained from Aldrich (Milwaukee, WI). RESULTS AND DISCUSSION Quasi-Molecular Ion Yield from Acidic Peptides. Figure 1 shows a comparison of gastrin I (1-14) negative- and positiveion MALDI-TOF spectra. From the same sample spot and with a common acidic matrix solution (pH ∼2), gastrin I (1-14) yielded a much more intense [M - H]- than its protonated counterparts. In addition, instead of the [M + H]+, the positive MALDI-TOF spectrum yielded molecular ion clusters consisting low-intensity alkaline metal adducts. Removing TFA from the matrix solution failed to increase the negative-ion yield. Surprisingly, addition of a base such as NH3 (pH ∼7) decreased negative-ion production. This illustrates that analysis of peptides in the negative mode does not require sample preparation methods different from the common procedures for analyzing positive ions. An experiment in the negative mode is therefore as simple as switching the polarity of the instrument’s voltages. To determine whether the cause of higher negative-ion intensity in gastrin I (1-14) is the array of the acidic residues (Glu), the same experiment was performed on two other peptides with several consecutive acidic residues: epidermal mitosis inhibiting pentapeptide and a synthetic enteropeptidase substrate hexapeptide. The pentapeptide yielded only [M - H]- while the hexapeptide yielded more intense [M + H]+ than [M - H]-. Of these three peptides, only the hexapeptide contains a highly basic residue (i.e., lysine, Lys). This result suggests that the intense [M - H]- of gastrin I (1-14) and the epidermal mitosis inhibiting pentapeptide is primarily due to their absence of highly basic residues (e.g., arginine, lysine, histidine) rather than the number of acidic residues. It should be noted that not all of the peptides studied yielded similar results. For example, hirudin (54-65) and Cys(Bzl)84CD4 (81-92) gave [M - H]- and [M + H]+ with similar intensities. Fibrinopeptide A and B give more intense [M + H]+ than [M - H]- by about a 2 to 1 ratio. Fragmentation of Deprotonated Pseudomolecular Ions. The nomenclature for fragment ions used in the following 5124 Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

Figure 3. PSD spectra of hirudin (54-65) in the (a) negative and (b) positive modes.

discussion is based on that proposed by Roepstorff and Fohlman.26 We have modified it for use in the negative-ion mode.22 Our symbols refer only to cleavage sites with additional hydrogens being specifically noted. For example, yn- is the C-terminal ion from the cleavage at the peptide bond (with no added hydrogens) while its positive counterpart is [yn + 2H]+. Figures 2 and 3 show the spectra of gastrin I (1-14) and hirudin (54-65), respectively, as representatives of the PSD spectra obtained in this study. Generally, the peaks from backbone cleavage are intense. At the lower mass region, the spectra contain weak peaks that were sometimes difficult to assign. As shown for hirudin (54-65) (Figure 3), some internal ions are intense. Overall the PSD spectra obtained in this study were simple and of high quality. Gastrin I (1-14). Gastrin I (1-14) is a fragment of the human gastrointestinal hormone gastrin I that is responsible for secretion of HCl and pepsinogen.10 The negative-ion PSD spectrum of gastrin I (1-14) is presented in Figure 2. Using the same sample preparation method, the intensity of [M + H]+ was too low for a PSD experiment. In addition, the pseudomolecular ions in the positive mode were adducts, e.g., [M + Na]+ and [M + K]+ (Figure 1). Therefore, a positive PSD spectrum would have been complicated by metal-mediated fragmentation and PSD fragments containing the various metal ions. (26) Roepstorff, P.; Fohlman, J. Biomed. Mass Spectrom. 1984, 11, 601.

Table 2. Fragmentation of Epidermal Mitosis Inhibiting Pentapeptidea n negative ion bncnx(6-n)[y(6-n) - 18]internal ionc

pEb 1

E 2 m s m m

D 3

m m

S 4

G 5

Table 3. Fragmentation of GDDDK-β-NPa,b n

positive iond a

Intensities where w is weak (60%). b pE, pyroglutamic acid. c The internal ion is composed of the underlined residues. d The [M + H]+ intensity was too low to be studied by PSD.

The PSD spectrum was dominated by an intense [yn - 18]series. Loss of H2O is probably from the carboxyl functional group on a Glu residue. Water loss from the y-ion series due to cleavage at Glu has been observed in CID of smaller peptides.14,17,18 The proposed fragmentation mechanism involves intramolecular proton transfer and nucleophilic attack on the carbonyl carbon. For analytical purposes, negative-ion PSD on gastrin I (1-14), with its five consecutive acidic residues, is an excellent method for sequence analysis. This is because the peptide yields primarily one type of ion, [yn -18]-, thus allowing instant amino acid information. The complete [yn - 18]- series is observed with the exception of n ) 1 and 14. Epidermal Mitosis Inhibiting Pentapeptide. The PSD fragment ions observed for this peptide, which inhibits mitosis of skin cells in mice,27 are listed in Table 2. Although C-terminal backbone ions and some internal ions form, the most intense ion is c2-, which involves cleavage between the two acidic residues Glu2-Asp3. However, cleavages adjacent to a single acidic residue are not always intense. The [y4 - 18]- (cleavage at Glp1-Glu2) and [y2 - 18]- (Asp3-Ser4) are intense while [y3 - 18]- (Glu2Asp3) is relatively weak. In addition to the [yn - 18]- series, x3and x4- are present. These are the only xn series ions observed in this study. Because this peptide did not yield [M + H]+, positive-ion PSD could not be performed. Synthetic Enteropeptidase Substrate. This synthetic peptide is used in fluorescence assays of enteropeptidase.28,29 Table 3 contains the PSD fragments for [M - H]- and [M + H]+ from this peptide. The negative-ion PSD spectrum is dominated by sequential loss of H2O from the y-series, such as [y2 - 36]-, [y4 - 36]-, and [y3 - 18]-, with the former being most intense. Internal ions [GD - 18]- and [DD - 18]- are also relatively intense. In the positive PSD spectrum, bn+ (n ) 2-6) were produced in moderate intensity and [yn + 2H]+ (n ) 1-4) with lower intensity. In addition, the spectrum showed [DK]+, [DDD]+, and (27) Reichelt, K. L.; Elgjo, K.; Edminson, P. D. Biochem. Biophys. Res. Commun. 1987, 146, 1493-1501. (28) Antonowicz, I.; Hesford, F. J.; Green, J. R.; Grogg, P.; Hadorn, B. Clin. Chim. 1980, 101, 69-76. (29) Hesford, F.; Hadorn, B.; Blaser, K.; Schneider, C. H. FEBS Lett. 1976, 71, 279-282.

D 2

negative ion [y(7-n) - 18)][y(7-n) - 36)]internal ionc

D 3

D 4

D 5

w s

w s

w s

m, GD-18

m

m, EDSG-18

G 1

positive ion bn+ [y(7-n)]+ internal ionc

m

K 6

m, DD-18

m w

s w

s w

m w

m, DDD a Intensities where w is weak (60%). b β-NP ) β-NH-C10H7. c The internal ion is composed of the underlined residues.

[K]+ internal ions. Although this peptide contains four acidic residues and only one basic residue, positive-ion PSD yielded more sequence information than negative-ion PSD. Hirudin (54-65). Hirudin (54-65) is a 14-residue Cterminal fragment of a protein with anticoagulating activity secreted by the leech Hirudo medicinalis.10 The peptide contains five acidic residues with two sets of two consecutive Glu and one isolated Asp. It yielded equally intense [M + H]+ and [M - H]-, as well as a simple and informative negative-ion PSD spectrum. The spectrum contains backbone cleavages at almost every residue (Figure 3a), with yn- and [yn - 18]- dominating. When Glu, is at the N-terminal end of a y-cleavage, H2O loss usually occurs; this leads to [yn - 18]-, n ) 4, 5, 8, and 9. This suggests that the Glu-Xxx is involved in [yn - 18]- formation (where Xxx represents any amino acid residue). In addition to intense backbone fragments, a number of internal ions were found. These are from two cleavages adjacent to acidic residues, such as [DFE]-, [PE]-, [EEIPE]-, [DFEEI]-, and [GDFEEIPE]-. The side-chain loss [y10 - 92]- and [M - H 92]- are unusual and have not been reported previously. The loss of 92 probably involves elimination of C6H4O from the tyrosine residue. For comparison, a positive-ion PSD spectrum of hirudin is presented in Figure 3b. Cleavage occurs at almost every residue on the peptide chain to yield bn+ ions, although b2+ and b7+ were missing. However, the complete amino acid sequence can be obtained by using a combined approach employing both positiveand negative-ion spectra. Fibrinopeptide A. Fibrinopeptide A and fibrinopeptide B (discussed below) are N-terminal peptides of fibrinogen AR and Bβ chains, respectively;10 they are released upon the conversion of fibrinogen to fibrin in blood clotting. Fibrinopeptide A contains three isolated acidic residues. Its negative PSD spectrum contains many backbone cleavages to yield yn- (Table 4). Among them, y10- (cleavage at Gly6-Asp7) and y6- (Ala10-Glu11) are the most intense. The remainder of the yn- series is relatively weak, but the amino acid sequence can still be determined. The spectrum contains some N-terminal ions such as a15(cleavage at Val15-Arg16), b13- (Gly13-Gly14), and b14- (Gly14-Gly15), which are from cleavages at basic or hydrophobic residues (Arg, Val, and Gly). Although the mechanistic detail requires further Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

5125

Table 4. Fragmentation of Fibrinopeptide Aa n

A 1

D 2

negative ion bnany(17-n)[y(17-n) - 18][y(17-n) - 36]internal ionb

S 3

G 4

E 5

G 6

D 7

F 8

L 9

A 10

E 11

G 12

G 13

G 14

w

w

V 15

R 16

m m

w

w m s

w

w

w s

m, ADSGE-36

m, EGGGVR-44

m, ADSGE-18

m, FLAEGG-44 w, SGEGDF

w, AE-18

w, VR-17

m, DFLAEGGGV-18 w, DFLAEGGGV positive ionc bn+ an+ y(17-n)+ [y(17-n) - 18]+

wd w

s

me

w m

w m

m w s

m w

w m

s

md s

w m we

m

a Intensities where w is weak (60%). b The internal ion is composed of the underlined residues. c No internal ions were found in the positive mode. d b3- and y2-ions have identical m/z of 274.3. e b4- and y3-ions have identical m/z of 331.3.

study, this process may be initiated by the deprotonated C-terminal group as proposed by Bowie14 and Beauchamp18 and their coworkers. In addition to backbone cleavages, fibrinopeptide A yields several internal ions, which are usually produced from cleavages adjacent to acidic residues. Examples include [SGEGDF]- (cleavages at Asp2-Ser3 and Phe8-Leu9) and [FLAEGG-44]- (Asp7-Phe8 and Gly13-Gly14). Generally, each fragment ion is intense, thus allowing facile peak assignment. Similar to negative-ion PSD, positive-ion fragmentation yielded several [yn + 2H]+ with some bn+ and an+. Cleavages at the acidic residues (n ) 5, 10, 11) to produce [yn + 2H]+ are intense. Although 10 members of the [yn + 2H]+ series are observed, five others in the series are missing. Therefore, to obtain the complete sequence, the PSD results from both positive and negative ions must be used. Fibrinopeptide B. Fragmentation from fibrinopeptide B, which contains one isolated Asp and two consecutive Glu residues, is summarized in Table 5. We have previously reported on the use of this peptide as a calibrant for negative-ion PSD.23 Most of the fragment ions are of the type yn-, n ) 3-10. For y-cleavages with an acidic residue on the N-terminal side, [yn - 18]- are more intense than yn-; for example, [y7 - 18]- > y7- and [y10 - 18]- > y10-. In addition to C-terminal fragments, N-terminal ions such as b13- and c13- (cleavage at Ala13-Arg14), c3- (Val3-Asn4), and c4- (Asn4-Asp5) were found. This differs from the high-energy CID spectra of Ser-containing peptides17 where y-ions corresponding cleavage at Ser and an extensive side-chain cleavages (CH2O) of Ser were found. In our PSD study of fibrinopeptide B, we observed neither yn- nor side-chain cleavage. Moreover, although asparagine (Asn) has induced backbone cleavage to yield yn- in a CID study,17 intense cn- corresponding to the cleavage adjacent to Asn have not previously been reported. 5126 Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

Finally, it should be noted that fibrinopeptides A and B contain a terminal basic residue, Arg. Both peptides fragment adjacent to this Arg, producing c(m-1)- where m is the position of Arg. A similar series of c-ions has been observed in the negative-ion PSD spectrum of a custom-synthesized peptide, (K2G4)2, containing the basic residue Lys.23 These observations suggest that the formation of the cn- may be promoted by the presence of Arg or Lys; this differs from cn+ formation where prerequisite sequences include an amino acid with OH functional group or a bulky side chain (Thr > Trp > Lys > Ser).30 Cys(Bzl)84-CD4 (81-92). Benzylated Cys(Bzl)84-CD4 (8192) fragment, whose PSD results are given in Table 6, has been found to inhibit human immunodeficiency virus (HIV) infection and the cell fusion process.31,32 The peptide contains five acidic residues with one isolated Glu and two sets of consecutive acidic residues (Glu7-Asp8 and Glu11-Glu12). The negative-ion PSD spectrum is dominated by many internal ions from cleavages at acidic residues; some are followed by the loss of H2O. The few backbone ions are intense and are almost completely limited to cleavage at the acidic residues Glu7-Asp8. For example, c7- from the cleavage at the Glu7-Asp8 is strong. This is our first observation of cn- from the cleavage adjacent to an acidic residue; normally only yn- is observed. However, c7- with the loss of 123 (SCH2C6H5) from the benzylated cysteine side chain is the most intense peak in the spectrum. Moreover, other ions with the same side-chain loss from Cys4 are also observed such as [M - H 123]- and [M - H - 123 - 18]-. In addition, loss of CH2C6H5 (-91) is abundant; for example, the [IC(Bzl)EV - 91]- internal (30) Downard, K. M.; Biemann, K. J. Am. Soc. Mass Spectrom. 1993, 4, 874881. (31) Nara, P. L.; Hwang, K. M.; Rausch, D. M.; Lifso, J. D.; Eiden, L. E. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 7139-7143. (32) Lifson, J. D.; Hwang, K. M.; Nara, P. L.; Fraser, B.; Padgett, M.; Dunlop, N. M.; Eiden, L. E. Science 1988, 241, 712-716.

Table 5. Fragmentation of Fibrinopeptide Ba n

pEb 1

G 2

negative ion bncny(15-n)[y(15-n) - 18][y(15-n) - 36]internal ionc

V 3

N 4

s

s

D 5

N 6

E 7

E 8

m w

w m s

G 9

F 10

F 11

S 12

A 13

w w

m m

m s

w

w m

w

R 14

m m w

w w

w, EG-18

m, AR

w, EG w, VN

w, NE

w, GFFS-18 m, EEGFFSA w, EEGFFS-44

positive ion bn+ y(15-n)+ [y(15-n) - 18]+ internal ionc

m

m

m

m w

m s

m w

m

m

m

m

m

m

w s

w m, GVNDNEEGFFSA

a Intensities where w is weak (60%). b pE, pyroglutamic acid. c The internal ion is composed of the underlined residues.

Table 6. Fragmentation of Cys(Bzl)84-Cd4a n negative ion bncn[cn- 123]- c [cn-123 - 18]- c y(13-n)[y(13-n) - 18][y(13-n) - 36]internal iond

T 1

Y 2

I 3

Cxb 4

E 5

V 6

E 7

D 8

Q 9

K 10

E 11

E 12

w m s m m w w

w m m m, M-123c m, M-123-18c

m, EVE w, TYICxE-18b

m, EDQK-18

m, ICxEV-91b

w, DQK w, EV

m, TYICxEVEDb w, EDQKEE positive ione bn+ yn+

s w

m m

m m

m m

s

m

m

m

m

a Intensities where w is weak (60%). b Cx ) Cys-Bzl, Cys-S-CH -C H . c 123 2 6 5 ) S-CH2-C6H5. d The internal ion is composed of the underlined residues. e No internal ions were found in the positive mode.

ion. Because this peptide produced several internal ions and sidechain losses, spectral interpretation was difficult. In contrast, the positive-ion PSD spectrum yields an almost complete series of [yn + 2H]+. The exceptions are n ) 1 and 2, which are adjacent to Glu residues. These y-ions allow easy determination of amino acid sequence. Positive-ion CID of alkylated Cys-containing peptides revealed extensive side-chain

loss33,34 due to cleavages of two bonds: Pep-SR and PepS-R. (Pep represents the peptide backbone, R represents an alkyl group, and the breaking bond is explicitly shown with a hyphen.) In a (33) Wolf, S. M.; Biemann, K. Int. J. Mass Spectrom. Ion Processes 1997, 160, 317-329. (34) Medzihradszky, K. F.; Maltby, D. A.; Qiu, Y.; Yu, Z.; Hall, S. C.; Chen, Y.; Burlingame, A. L. Int. J. Mass Spectrom. Ion Processes 1997, 160, 357-369.

Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

5127

SIMS/high-energy CID work, Wolf and Biemann33 observed both Pep-SR and PepS-R cleavages; however, only PepS-R cleavage was reported in a MALDI-CID study by Medzihradszky et al.34 Although both this latter study and our work employed MALDI, we observed only cleavage of the Pep-SR bond in positive-ion PSD of Cys(Bzl)84-CD4 (81-92). CONCLUSION The negative-ion MALDI-TOF PSD experiment has been shown to be useful in analyzing acidic peptides. Overall, the negative-ion PSD spectra contain backbone fragment ions that allow the determination of the peptide sequence. The most common backbone series is yn-, which is frequently accompanied by H2O loss. Also, as supported by the fibrinopeptide A and fibrinopeptide B data, useful results from negative-ion PSD are not limited to highly acidic peptides. Different patterns of the acidic residues and of amino acid composition result in various fragmentations and molecular ion yields. For example, an array of acidic residues, as well as the absence of basic residues, increases the ion yield of the [M H]- for gastrin I (1-14) and epidermal mitosis inhibiting pentapeptide. In addition, the spectra of the two peptides showed abundant fragment ions with an almost complete series of [yn 18]-. Compared with high-energy CID studies on smaller deprotonated peptides, under PSD conditions, larger peptide anions did not show extensive side-chain fragmentation. The lone exceptions

5128 Analytical Chemistry, Vol. 70, No. 24, December 15, 1998

are the loss of large side chains such as C6H4O from Tyr and SCH2C6H5 from Cys(Bzl) in hirudin (54-65) and Cys(Bzl)84-CD4 (81-92), respectively. This observation may be due to differences in the fragmentation mechanisms and energy deposition between CID and metastable fragmentation. The size of the peptides may also be a factor. Although positive-ion PSD spectra for acidic peptides showed backbone cleavages that facilitate peptide sequencing, the complete sequence of ions of the same type (e.g., bn+ or yn+) is often not observed. Generally, a combined approach using both the negative and positive PSD spectra can facilitate peptide sequencing. ACKNOWLEDGMENT Funding to purchase the mass spectrometer was provided by the National Science Foundation Academic Research Infrastructure Program (Grant CHE-9413529) and the Ohio Board of Regents Action Fund. In addition, support from the Ohio Board of Regents Research Challenge Program is gratefully acknowledged. J.J. thanks the Miami University Graduate School for the Dissertation Scholarship for the academic year 1997-1998.

Received for review May 27, 1998. Accepted September 23, 1998. AC980577N