Simplifying fragmentation patterns of multiply charged peptides by N-terminal derivatization and electron transfer collision activated dissociation

Article abstract of DOI:10.1021/ac9000942

N-terminal peptide derivatization strategies used in conjunction with tandem mass spectrometry to yield simplified fragmentation patterns have shown limited success for the de novosequencing of multiply charged peptides, including those predominantly form

Full text of DOI:10.1021/ac9000942

Anal. Chem. 2009, 81, 3645–3653
Simplifying Fragmentation Patterns of Multiply
Charged Peptides by N-Terminal Derivatization and
Electron Transfer Collision Activated Dissociation
James A. Madsen and Jennifer S. Brodbelt*
Department of Chemistry and Biochemistry, The University of Texas at Austin, 1 University Station A5300,
Austin, Texas 78712
N-terminal peptide derivatization strategies used in con-
junction with tandem mass spectrometry to yield simpli-
fied fragmentation patterns have shown limited success
for the de novo sequencing of multiply charged peptides,
including those predominantly formed in LC-ESI-MS
experiments. Significant proton mobilization occurs for
multiply charged peptides upon collisional activation,
resulting in the formation of both N-terminal and C-
terminal product ions rather than an exclusive series of
C-terminal ions preferred for de novo sequencing algo-
rithms. To circumvent this problem, multiply charged,
N-terminally derivatized peptides were subjected to elec-
tron transfer reactions with fluoranthene anions to pro-
duce singly charged, radical species. Upon subsequent
“soft” collision induced dissociation (CID), highly abun-
dant z-type ions were formed nearly exclusively, which
yielded simplified fragmentation patterns amenable to de
novo sequencing methods. Furthermore, the derivatized
peptides retained labile phosphoric acid moieties, and the
enhanced set of z ions were also observed for peptides
not possessing basic C-terminal residues, a type of peptide
that poses more challenges to traditional simplification
methods based on collision activated dissociation. This
improved LC-MSn strategy was demonstrated for a
variety of multiply charged model peptides and a tryptic
digest of myoglobin.
tandem MS is performed on the resulting peptides. De novo
algorithms or cross-correlation scoring methods can then be used
to decipher the tandem mass spectra, assign peptide sequences,
and integrate the data sets to identify proteins.2-8 However, often
the tandem mass spectra are extremely complex with many
redundantproductionsthatcauseerrorsinspectralinterpretation,9-11
leading to incorrect identification of proteins or unidentified
proteins. To enhance the success of de novo sequencing methods,
measures have been taken to reduce the complexity of product
ion spectra. Keough et al. first reported the derivatization of the
N-terminus of peptides with sulfonic acid groups, which upon
dissociation yield mass spectra with an extensive array of C-
terminal ions (e.g., y ions) void of neutralized N-terminal ions (e.g.,
b ions).12 This technique greatly improved the accuracy of de novo
peptide sequencing methods. A number of other studies have
described the potential merits of other derivatization procedures
for proteomic applications.10,11,13-34 For example, the N-terminal
sulfonation reagent 4-sulfophenyl isothiocyanate (SPITC) in con-
(2) Eng, J. K.; McCormack, A. L.; Yates, J. R., III J. Am. Soc. Mass Spectrom.
1994, 5, 976–989
(3) Perkins, D. N.; Pappin, D. J. C.; Creasy, D. M.; Cottrell, J. S. Electrophoresis
1999, 20, 3551–3567
(4) Taylor, J. A.; Johnson, R. S. Anal. Chem. 2001, 73, 2594–2604
(5) Johnson, R. S.; Taylor, J. A. Mol. Biotechnol. 2002, 22, 301–315
(6) Taylor, J. A.; Johnson, R. S. Rapid Commun. Mass Spectrom. 1997, 11,
1067–1075
(7) Dancik, V.; Addona, T. A.; Clauser, K. R.; Vath, J. E.; Pevzner, P. A.
J. Comput. Biol. 1999, 6, 327–342
(8) Ma, B.; Zhang, K.; Hendrie, C.; Liang, C.; Li, M.; Doherty-Kirby, A.; Lajoie,
G. Rapid Commun. Mass Spectrom. 2003, 17, 2337–2342
(9) Mann, M.; Wilm, M. Anal. Chem. 1994, 66, 4390–4399
(10) Sergeant, K.; Samyn, B.; Debyser, G.; Van Beeumen, J. Proteomics 2005,
5, 2369–2380
(11) Wilson, J. J.; Brodbelt, J. S. Anal. Chem. 2006, 78, 6855–6862
(12) Keough, T.; Youngquist, R. S.; Lacey, M. P. Proc. Natl. Acad. Sci. U.S.A.
1999, 96, 7131–7136
(13) Marekov, L. N.; Steinert, P. M. J. Mass Spectrom. 2003, 38, 373–377
(14) Lee, Y. H.; Kim, M.-S.; Choie, W.-S.; Min, H.-K.; Lee, S.-W. Proteomics 2004,
4, 1684–1694
(15) Lee, Y. H.; Han, H.; Chang, S.-B.; Lee, S.-W. Rapid Commun. Mass Spectrom.
2004, 18, 3019–3027
(16) Shin, J.-W.; Lee, Y. H.; Hwang, S.; Lee, S.-W. J. Mass Spectrom. 2007, 42,
380–388
(17) Samyn, B.; Debyser, G.; Sergeant, K.; Devreese, B.; Van Beeumen, J. J. Am.
Soc. Mass Spectrom. 2004, 15, 1838–1852
(18) Wang, D.; Kalume, D.; Pickart, C.; Pandey, A.; Cotter, R. J. Anal. Chem.
2006, 78, 3681–3687
(19) Wang, D.; Cotter, R. J. Anal. Chem. 2005, 77, 1458–1466
(20) Keough, T.; Lacey, M. P.; Youngquist, R. S. Rapid Commun. Mass Spectrom.
2000, 14, 2348–2356
(21) Wang, D.; Kalb Suzanne, R.; Cotter, R. J. Rapid Commun. Mass Spectrom.
2004, 18, 96–102
.
.
.
.
.
.
.
Mass spectrometry (MS) has increasingly become the analyti-
cal method of choice in the field of proteomics in recent years
due to significant advances in instrumentation, sampling tech-
niques, and data interpretation algorithms. Although there have
been tremendous inroads, there also remain some unresolved
limitations in practical applications. For example in a recent study,
24 proteomics laboratories were given a sample of Escherichia
coli spiked with 20 proteins and of these participating laboratories,
only six correctly identified the 20 proteins.1 While discouraging,
these results were a significant improvement compared to a similar
study conducted the previous year. One key problem resulting
in protein misidentification is a high false positive rate in computer-
guided interpretation of tandem mass spectra. In a typical “bottom-
up” approach, proteins are first enzymatically digested, and then
.
.
.
.
.
.
.
.
.
.
.
* To whom correspondence should be addressed. E-mail: jbrodbelt@
mail.utexas.edu.
.
(1) Service, R. F. Science 2008, 321, 1758–1761
.
.
10.1021/ac9000942 CCC: $40.75  2009 American Chemical Society
Published on Web 03/27/2009
Analytical Chemistry, Vol. 81, No. 9, May 1, 2009 3645

junction with ESI-MS/MS and MALDI-MS/MS has proved to be
useful identifying sites of ubiquitination.18,19 Our group has used
N-terminal sulfonation with infrared multiphoton dissociation
(IRMPD) to improve de novo sequencing of peptides by eliminat-
ing the low mass cutoff inherent to ion traps.11 Recently, we have
synthesized a new N-terminal reagent with a highly IR-absorbing
phosphonite group that increased the dissociation efficiencies and
sequence information obtained upon IRMPD.35
to be particularly promising for characterization of post-transla-
tional modifications (PTMs)38-45 and for increasing peptide
sequence coverage.36,46 Both methods involve the reaction of
multiply charged cations with either low-energy electrons (ECD)
or radical anions (ETD), producing complementary c- and z-type
backbone cleavages that retain labile PTMs such as phosphory-
lation. For peptides of lower charge, however, these electron-based
methods yield charge-reduced radicals from the precursor as the
most abundant products with low yields of c and z products.36,47
Recently, both the Coon and McLuckey groups have implemented
supplemental collisional activation (i.e., low-energy CID) of charge-
reduced peptide radicals after the electron transfer reaction
(ETcaD), thus affording significantly improved peptide sequence
coverage.47-49 Both ETD and ECD are promising new alternatives
to CID for sequencing peptides, but the resulting spectra remain
rather complex with both N-terminus and C-terminus product ions
that can complicate de novo interpretation.
In this study, doubly protonated peptides derivatized at the
N-terminus via reactions with 4-sulfophenyl isothiocyanate (SPITC)
or 4-(chlorosulfonyl)phenyl isocyanate (SPC) (Supplemental Fig-
ure 1) were converted to charged-reduced radical species after
gas-phase electron transfer (ET) reactions with fluoranthene
radicals. The resulting singly charged radical species were then
subjected to subsequent collisional activation (ETcaD), which
produced an enhanced series of z ions void of N-terminal ions.
This spectral simplification strategy was performed using a linear
ion trap on a liquid chromatographic time scale, and improved de
novo sequencing was realized for a variety of multiply charged
peptides including phosphorylated species.
One major drawback to the use of N-terminal derivatization
strategies is that the simplification of the fragmentation patterns
of peptides is substantially impaired when multiply charged
precursor ions are analyzed.14,15,18,20 For example, when doubly
charged SPITC-modified peptides are dissociated, the N-terminal
b ions are no longer neutralized due to the presence of the extra
mobile proton, and the utility of the derivatization procedure is
greatly reduced.14 This is particularly problematic for LC-ESI-
MS applications in which tryptic peptides are typically multiply
protonated (i.e., charge states of 2+ and greater).36 Lee et al.
described a method to simplify the fragmentation patterns of these
multiply charged ions by using isotopically labeled SPITC, but
extra derivatization steps and more elaborate spectral interpreta-
tion were needed.15 Our group has attached UV chromophores
to the N-terminus of peptides, and the modified peptides almost
exclusively produced y ions upon ultraviolet photodissociation
(UVPD), a striking selectivity attributed to the secondary dis-
sociation and rapid annihilation of the UV chromophore-containing
b ions.34 However, because of the slow repetition rate of the laser
(10 Hz) and the requirement for multiple laser pulses used in this
application, analysis on a chromatographic time scale could not
be achieved.
EXPERIMENTAL SECTION
In recent years, other ion activation methods have been
developed as alternatives to collision induced dissociation (CID,
also known as CAD). Electron-based dissociation techniques such
as electron capture dissociation (ECD)37 in FTICR instruments
and electron transfer dissociation (ETD)38 in ion traps have shown
Materials. Myoglobin and cytochrome c from equine heart
and all reagents were purchased from Sigma-Aldrich (St. Louis,
MO). The peptides KRPPGFSPFR and ASHLGLAR were obtained
from BACHEM (King of Prussia, PA), the phosphorylated peptides
KRpTIRR and TRDIYETDYpYRK from AnaSpec (San Jose, CA),
and immobilized TPCK-treated trypsin beads from Pierce Bio-
technology, Inc. (Rockford, IL). All solvents were purchased from
Fisher Scientific (Fairlawn, NJ).
(22) Wang, D.; Xu, W.; McGrath, S. C.; Patterson, C.; Neckers, L.; Cotter, R. J.
J. Proteome Res. 2005, 4, 1554–1560
.
(23) Pashkova, A.; Chen, H.-S.; Rejtar, T.; Zang, X.; Giese, R.; Andreev, V.;
Moskovets, E.; Karger, B. L. Anal. Chem. 2005, 77, 2085–2096.
(24) Oehlers, L. P.; Perez, A. N.; Walter, R. B. Rapid Commun. Mass Spectrom.
2005, 19, 752–758.
(47) Swaney, D. L.; McAlister, G. C.; Wirtala, M.; Schwartz, J. C.; Syka, J. E. P.;
Coon, J. J. Anal. Chem. 2007, 79, 477–485
(37) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998,
120, 3265–3266
(38) Syka, J. E. P.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt, D. F.
Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528–9533
(39) Mirgorodskaya, E.; Roepstorff, P.; Zubarev, R. A. Anal. Chem. 1999, 71,
4431–4436
(40) Stensballe, A.; Jensen, O. N.; Olsen, J. V.; Haselmann, K. F.; Zubarev, R. A.
Rapid Commun. Mass Spectrom. 2000, 14, 1793–1800
(41) Hakansson, K.; Cooper, H. J.; Emmett, M. R.; Costello, C. E.; Marshall,
A. G.; Nilsson, C. L. Anal. Chem. 2001, 73, 4530–4536
(42) Kleinnijenhuis, A. J.; Kjeldsen, F.; Kallipolitis, B.; Haselmann, K. F.; Jensen,
O. N. Anal. Chem. 2007, 79, 7450–7456
.
(25) Guillaume, E.; Panchaud, A.; Affolter, M.; Desvergnes, V.; Kussmann, M.
Proteomics 2006, 6, 2338–2349.
(26) Bauer, M. D.; Sun, Y.; Keough, T.; Lacey, M. P. Rapid Commun. Mass
Spectrom. 2000, 14, 924–929.
(27) Bao, J.; Ai, H.; Fu, H.; Jiang, Y.; Zhao, Y.; Huang, C. J. Mass Spectrom. 2005,
40, 772–776.
(28) Wang, F.; Fu, H.; Jiang, Y.; Zhao, Y. J. Am. Soc. Mass Spectrom. 2006, 17,
995–999.
(29) Beardsley, R. L.; Reilly, J. P. J. Am. Soc. Mass Spectrom. 2004, 15, 158–
167.
(30) Beardsley, R. L.; Sharon, L. A.; Reilly, J. P. Anal. Chem. 2005, 77, 6300–
6309.
(31) Brancia, F. L.; Butt, A.; Beynon, R. J.; Hubbard, S. J.; Gaskell, S. J.; Oliver,
S. G. Electrophoresis 2001, 22, 552–559.
.
.
.
.
.
.
(43) Srikanth, R.; Wilson, J.; Bridgewater, J. D.; Numbers, J. R.; Lim, J.; Olbris,
M. R.; Kettani, A.; Vachet, R. W. J. Am. Soc. Mass Spectrom. 2007, 18,
(32) Summerfield, S. G.; Bolgar, M. S.; Gaskell, S. J. J. Mass Spectrom. 1997,
32, 225–231.
1499–1506
(44) Zabrouskov, V.; Ge, Y.; Schwartz, J.; Walker, J. W. Mol. Cell. Proteomics
2008, 7, 1838–1849
(45) Sweet, S. M. M.; Mardakheh, F. K.; Ryan, K. J. P.; Langton, A. J.; Heath,
J. K.; Cooper, H. J. Anal. Chem. 2008, 80, 6650–6657
(46) Molina, H.; Matthiesen, R.; Kandasamy, K.; Pandey, A. Anal. Chem. 2008,
80, 4825–4835
(48) Han, H.; Xia, Y.; McLuckey, S. A. Rapid Commun. Mass Spectrom. 2007,
21, 1567–1573
(49) Xia, Y.; Han, H.; McLuckey, S. A. Anal. Chem. 2008, 80, 1111–1117
.
(33) Yamaguchi, M.; Oka, M.; Nishida, K.; Ishida, M.; Hamazaki, A.; Kuyama,
H.; Ando, E.; Okamura, T.-a.; Ueyama, N.; Norioka, S.; Nishimura, O.;
Tsunasawa, S.; Nakazawa, T. Anal. Chem. 2006, 78, 7861–7869.
(34) Wilson, J. J.; Brodbelt, J. S. Anal. Chem. 2007, 79, 7883–7892.
(35) Vasicek, L. A.; Wison, J. J.; Brodbelt, J. S. J. Am. Soc. Mass Spectrom., 2009,
20, 377-384.
.
.
.
(36) Kjeldsen, F.; Giessing, A. M. B.; Ingrell, C. R.; Jensen, O. N. Anal. Chem.
.
2007, 79, 9243–9252
.
.
3646 Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 1. CID mass spectra of the doubly charged peptide ASHLGLAR (a) derivatized with SPITC, 12 mV collision voltage, q ) 0.25; (b)
derivatized with SPC, 14 mV collision voltage, q ) 0.25. The unmodified peptide is denoted by M and the precursor ion by an asterisk (*). The
partial decomposition of the SPITC or SPC moiety resulting in a bonded carbon and sulfur atom attached N-terminally to the peptide is denoted
by CS.
Derivatization and Sample Preparation. Before N-terminal
derivatization, lysines were converted to homoarginine residues
through a guanidination step (except where noted) to ensure
modification with 4-sulfophenyl isothiocyanate (SPTIC) exclusively
at the N-terminus and not at the ε-amine group of lysines. A stock
of 0.05 g of O-methylisourea in 50 µL of water was prepared, and
5 µL of this stock solution, 50 µL of 5 N ammonium hydroxide,
and approximately 10 nmol of the protein digest or model peptide
was combined. This mixture was then incubated for 10 min at 65
°C, after which the sample was desalted via C18 spin columns.
The SPITC and 4-(chlorosulfonyl)phenyl isocyanate (SPC)
derivatization procedures used in this study were adapted from
previous reports.16,50 Briefly, a stock solution of SPITC was made
by dissolving approximately 1 mg of SPTIC in 100 µL of a 20 mM
NaHCO3 solution (pH ∼9.5). A 20 µL aliquot of this stock was
combined with 10 nmol of peptide or protein digest, which was
then incubated for approximately 30 min at 55 °C. For SPC-
modified peptides, a stock solution was made by dissolving
approximately 1 mg of SPC in 100 µL of acetonitrile. A 20 µL
aliquot of this stock was combined with 10 nmol of peptide in
a 50:50 pyridine/water (v/v) solution, which was then incubated
for approximately 30 min at ambient temperature. After incuba-
tion, all samples were cleaned up using C18 spin columns. For
direct infusion experiments, working solutions of 10 µM
peptides in 49.5:49.5:1.0 water/methanol/acetic acid (v/v/v)
solution were prepared before infusion at a flow rate of 3 µL/
min into the mass spectrometer.
Digestion occurred for 18 h at 37 °C, and the beads were discarded
after digestion.
Mass Spectrometry and Liquid Chromatography. Mass
spectrometric analysis was performed on a Thermo Fisher (San
Jose, CA) LTQ XL two-dimensional linear ion trap mass spec-
trometer equipped with an ETD unit. Fluoranthene anions were
introduced as the electron-transfer reagent for ETD experiments
with reaction times of 100 ms. Standard parameters (q-value equal
to 0.25 and activation times of 30 ms) were used for all CID
experiments. ETcaD was carried out using ET reaction times of
100 ms, followed by isolation of the charge-reduced radical, and
a 30 ms CID period. For all MSn experiments, a product ion was
considered to be detected in an ETD, CID, or ETcaD mass
spectrum if the ion peak had a signal/noise ratio equal to or
greater than 3.
A Hitachi L-7000 (Hitachi Ltd.) analytical HPLC system was
used for liquid chromatographic separations. Samples containing
10 µM digested protein were injected (10 µL) onto a Symmetry300
reversed-phase C18 column (Waters, Milford, MA) (2.1 mm ×
50 mm, 3.5 µm packing) with a matched guard column (2.1
mm × 10 mm, 3.5 µm packing). Eluents consisted of 0.2% formic
acid in water (A) and 0.2% formic acid in acetonitrile (B).
Gradient elution was performed as follows: 95% (A) for 2 min
followed by a linear gradient to 40% (A) over 60 min at a flow
rate of 0.300 mL/min. Data-dependent aquisition was used for
all LC-MSn analysis with the first scan event being a full mass
spectrum at a m/z range of 400 - 2000 and subsequent events
consisting of ETD, CID, and ETcaD of the one or two most
abundant peaks. A normalized collision energy of 35% (61-142
mV) was used for all LC-CID-MS analyses.
For protein samples, 100 µL of immobilized TPCK-treated
trypsin beads and 10 µmol of ammonium bicarbonate were used
to digest 10 nmol of protein prior to the derivatization procedures.
(50) Wang, D.; Kalb, S. R.; Cotter, R. J. Rapid Commun. Mass Spectrom. 2003,
18, 96–102
.
Analytical Chemistry, Vol. 81, No. 9, May 1, 2009 3647

RESULTS AND DISCUSSION
also seen in Figure 2a, product ions attributed to partial cleavage
of the SPITC moiety are observed in the m/z range of 850-1040.
These product ions are thought to stem from the low-energy,
radical-driven dissociation processes of ETcaD since many of these
ions were not observed upon CID of the singly charged, even-
electron peptides. Regardless, these ions do not cause substantial
spectral congestion as they do not overlap with the z ion series,
and these were formed in low abundance.
N-Terminally Derivatized Multiply Charged Peptides. Two
commonly used commercial reagents (e.g., SPITC and SPC) were
used for N-terminal derivatization reactions of the peptides,
resulting in a mass addition of 215 Da for SPITC and 199 Da for
SPC. The structures of both reagents, SPITC and SPC, are shown
in Figure 1 in the Supporting Information. As noted previously,
the chlorine atom in SPC is hydrolyzed and replaced with a
hydroxyl group in aqueous solution.16 The sulfonic acid group
present in both reagents introduces a negative charge at the
N-terminus, thus reducing the overall charge of the peptides and
simplifying the tandem mass spectra by effectively neutralizing
N-terminal product ions (e.g., b, a, and c ions). This simplification
of the product ion spectra of singly charged sulfonate-derivatized
peptides has been documented previously and has proven to be
especially beneficial for de novo sequencing strategies.10-12,26
Upon CID, an enhanced set of C-terminal ions (y ions) are
produced, and redundant N-terminal products (b, a ions) are
formed as neutral species and are not detected. However, when
more than one positive charge resides on the derivatized peptide,
the N-terminal products formed upon CID may retain a charge,
thus nulling the benefit of the derivatization process for spectral
simplification. This increased spectral convolution is illustrated
in Figure 1 for the doubly charged derivatized peptide ASHLGLAR
in which N- and C-terminal product ions of multiple charge states
are detected. Upon ESI, the N-terminal derivatized peptide
ASHLGLAR was predominantly observed in the 2+ charge state
(data not shown). CID of the doubly charged ions yielded complex
spectra containing both y and redundant a and b ions (with these
N-terminal ions retaining the original SPITC- or SPC-modification)
(see parts a and b of Figure 1). Moreover, product ions in multiple
charge states are observed, as is the case for the y7 ion, which
further complicates spectral interpretation. This loss in spectral
simplification for multiply charged peptides is the primary
reason that N-terminal sulfonation reactions have mainly been
employed for MALDI-MS, not ESI-MS, applications.
An alternative approach, as described in the present study,
relies on using ETD (via ETcaD) rather than CID for characteriza-
tion of the N-terminal derivatized peptides. In this case, doubly
charged SPITC- or SPC-modified peptides were subjected to
electron transfer reactions, producing low amounts of sequence
ions and predominantly charge-reduced peptide radical cations
(singly charged) as illustrated in Figure 2 in the Supporting
Information. Subsequently, the resulting charged-reduced ions
were subjected to collisional activation, a net process termed
ETcaD, which yielded a simple set of C-terminal (z ions) without
any redundant N-terminal ions (c, b, or a ions) (see Figure 2).
The ETcaD spectrum of doubly charged SPITC-modified
ASHLGLAR is illustrated in Figure 2a, displaying a complete series
of z-type ions and no N-terminal ions. Some z product ions were
observed as radicals and others as even electron ions, a result
which is discussed in more detail later. Also, a product ion
attributed to the loss of both the SPITC moiety and ammonia (net
loss of 232 Da) was observed in all ETcaD spectra; this charac-
teristic loss can be used as a marker for every SPITC-modified
peptide. It is a spectral feature that is especially convenient when
the derivatization of peptide mixtures might be incomplete, and
one needs a facile means to screen the species of interest. As
ETD of SPITC- or SPC-derivatized peptides with more than
two charges resulted in significant abundances of both c- and
z-ions and increased sequence coverage than obtained upon ETD
of the doubly charged SPITC- or SPC-derivatized peptides, which
is analogous to the ETD behavior of unmodified peptides (i.e.,
more efficient formation of both c- and z- ions with increasingly
higher precursor charge state). However, significant abundances
of nondissociated charge-reduced peptides were also observed.
Because the ETD spectra for the SPITC- and SPC-modified
peptides in the higher charge states display both the N-terminal
and C-terminal sequence ions, the resulting spectra contain
redundant ions that are less amenable to de novo sequencing.
Formation of zn-1 versus yn-1 Product Ions. It has
previously been shown that yn-1 is the most abundant product
formed upon CID of SPITC-modified peptides, both singly and
,15
doubly charged.14 This phenomenon is illustrated in Figure
1a in which both the singly and doubly charged y7 ions dominate
the CID mass spectrum obtained for doubly protonated SPITC-
ASHLGLAR. This process is promoted by nucleophilic attack
of the sulfur atom of the SPITC moiety on the carbonyl oxygen
atom of the adjacent amino acid, resulting in an Edman-type
,32
degradation.14 Derivatization of peptides with SPC, which is
identical to SPITC except for substitution of an oxygen atom for
the sulfur, alleviates the domination of the yn-1 ion due to the
reduced nucleophilicity of the oxygen.16 This leads to a more
even distribution of y ions that is preferable for de novo
sequencing (Figure 1b); however, both y and b series were
produced upon CID of the doubly protonated, SPC-derivatized
peptide which led to numerous redundant ions and a cluttered
spectrum. These CID patterns are characteristic of other doubly
charged, SPITC- and SPC-derivatized peptides: dominant formation
of the yn-1 ion for the former and more importantly cluttered
spectra for both; thus, neither was ideal for sequencing peptides
in mixtures using an LC-ESI-MSn approach.
In contrast, ETcaD of the SPITC-modified peptides using low
supplemental collision voltages yielded a broad series of z ions
and did not result in enhanced formation of the yn-1 (or
corresponding zn-1) product ions, a result that was surprising
given the lability of the peptide bond that normally is cleaved
to form the yn-1 ion upon CID (see Figure 2a with comparison
to Figure 1a). However, ETcaD using a higher collision voltage
produced a more abundant yn-1 ion, in addition to the series of
z ions (Figure 2b). The yn-1 ion is easily identifiable because it
is 16 Da greater in mass than the zn-1 ion, but its presence
can largely be eliminated by using lower energy ETcaD
conditions as demonstrated in Figure 2a. To optimize the
formation of the preferred z-type ions over the low energy yn-1
product, the collision voltage and q-value were varied systemati-
3648 Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 2. ETcaD mass spectra of the doubly charged peptide ASHLGLAR (a) derivatized with SPITC, 24 mV subsequent collision voltage, q
) 0.12; (b) derivatized with SPITC, 42 mV subsequent collision voltage, q ) 0.16; (c) derivatized with SPC, 42 mV subsequent collision voltage,
q ) 0.16. Product ions due to partial decomposition of the SPITC or SPC moiety are denoted by #. The unmodified peptide is denoted by M,
and the charge-reduced radical precursor ion is denoted by *.
cally during the CID activation period. Figure 3 illustrates the
resulting trend for formation of the zn-1 versus yn-1 ions for
doubly protonated SPITC-modified ASHLGLAR. As seen in
Figure 3, the zn-1 dominates at lower energy ETcaD conditions
and the yn-1 ion dominates at higher energy ETcaD conditions
(i.e., higher q-value and/or greater CID voltage). These
experiments parallel Coon’s initial ETcaD work where more b
and y ions (instead of c and z) were seen at higher q-values
and greater collision energies.47 The absence of any y product
ions using lower energy ETcaD conditions suggests that the
radically driven dissociation pathway that leads to z-type ions
could be a lower energy process compared to the process that
leads to the formation of the yn-1 ion.
conditions; however, some low abundance y ions (e.g., y4 and
y5) were detected at high q-values and CID energies. Because
the SPC derivatization procedure involves the use of a less
desirable 50% pyridine solution and reaction yields were often
low, SPITC was used as the N-terminal derivatizing reagent
for the remainder of the study. The SPITC reaction procedure
consistently produced the highest yields of modified peptides.
For example, guanidination and SPITC derivatization of the
tryptic digest of myoglobin (described in more detail in
subsequent sections) resulted in 100% conversion of unmodified
peptides to modified species.
LC-MSn Analysis of SPITC-Derivatized Tryptic Pep-
tides. To evaluate the analytical utility of the ETcaD method for
protein characterization, myoglobin was selected as a model
protein for tryptic digestion and SPITC derivatization. Under
LC-MS conditions, most of the resulting SPITC-modified tryptic
peptides were observed in the 2+ charge state upon ESI. Thus,
Another way to suppress the formation of the yn-1 ion is to
use SPC as the derivatization reagent. Figure 2c shows the
ETcaD mass spectrum of doubly charged SPC-modified
ASHLGLAR, and no yn-1 ion was formed using any ETcaD
Analytical Chemistry, Vol. 81, No. 9, May 1, 2009 3649

Figure 3. Percent abundance of zn-1 versus yn-1 (i.e., zn-1/(zn-1 + yn-1) × 100) in ETcaD spectra as a function of CID voltage and q-value for
doubly charged SPITC-ASHLGLAR.
for data-dependent LC-MS experiments, the doubly charged
peptides were typically the ones subjected to ion activation.
ETcaD, CID, and ETD spectra were collected in a data-dependent
manner for the tryptic peptides in several charge states for
separate LC runs of SPITC-derivatized and nonderivatized pep-
tides. The peak areas of the product ions for all C-terminal ions
and all N-terminal ions were summed separately and converted
to percentages. The results are displayed in Figure 4 to allow ready
comparison of the preference for formation of C-terminal (z ions)
versus N-terminal product ions and other redundant ions such as
ones in multiple charge states and redundant C-terminal species.
The extreme C-terminal product ion selectivity for the ETcaD
spectra obtained for the doubly charged SPITC-peptides is readily
apparent, as demonstrated by the percentages of C-terminal ions
ranging from 95.3% to 100.0% with an average of 97.1% in Figure
4. In comparison, CID of the doubly charged SPITC-modified
peptides produced more complicated spectra containing both
N-terminal and C-terminal ions. Three of the peptides (LFTGH-
PETLEK, FDKFK, and HLKTEAEMK) appeared to show high
C-terminal ion selectivity upon CID, but this was due to production
of very dominant yn-1 ions and not due to production of a
comprehensive series of y ions as would be needed for effective
de novo sequencing. Because of the limited m/z range of the
ion trap (up to 2000 m/z), larger tryptic peptides such as
YLEFISDAIIHVLHSK were unable to be analyzed by the
ETcaD method after N-terminal derivatization. Although ETD
alone and ETD with supplemental CID of unmodified peptides
has been shown to increase the level of sequence coverage
and also to promote retention of labile modifications in the
product ions,36,38,47 these electron transfer based methods yield
spectra that are often much more complicated than traditional CID
spectra because highly abundant c-, z-, and a-type ions are present
as well as low abundance y ions. Thus, the SPITC derivatization
procedure in conjunction with ETcaD, not ETD alone, is a key
factor that leads to the preferential formation of C-terminal ions.
Both ETcaD and CID can be set up as successive scan events
during a single data-dependent LC-MS run; therefore, these
methods can be used in a complementary rather than competitive
mode to increase the overall sequence coverage. In the present
study, the average sequence coverage obtained was 86% for all
N-terminally derivatized peptides analyzed by ETcaD (shown in
Figure 4) with two major factors that limited the sequence
coverage: (1) the low mass cutoff (LMCO) of the ion trap
instrument prohibited detection of the lowest m/z product ions
and (2) miscleavages from proline residues inherent to ETD. The
high sequence coverage obtained was comparable to previous
reports of 89% coverage of unmodified peptides by ETcaD.47 Also,
the increased spectral simplification afforded by ETcaD of the
SPITC-modified peptides could better unveil miscleavages stem-
ming from the presence of proline residues based on character-
istics gaps in the observed series of z ions. In summary, the
combination of ETcaD with SPITC-derivatization affords the best
outcome of high sequence coverage and most simplified spectra
containing nearly exclusively z ions.
Phosphorylated Peptides. One of the biggest advantages of
using electron-based dissociation techniques such as ETD is that
labile modifications of amino acids generally remain intact, a factor
that greatly aids in identifying PTMs.38-45 However, loss of
phosphoric acid is still a substantial dissociation pathway for
phosphopeptides examined by ETcaD in a linear ion trap.47 To
circumvent this shortcoming, we evaluated the use of ETcaD for
analysis of the N-terminal sulfonated phosphorylated peptide
KRpTIRR. The ETcaD spectra for the SPITC-derivatized peptides
(i.e., ET of the 2+ and 3+ charge states, then collisional activation
of the resulting singly charged radicals; see parts a and b of Figure
5, respectively) displayed an enhanced series of z ions with
minimal loss of phosphoric acid. Conversely, ETcaD of doubly
charged unmodified KRpTIRR (Figure 5c), CID of singly charged
SPITC-modified KRpTIRR (Figure 5d), and CID of doubly charged
SPITC-modified KRpTIRR (Figure 5e) all showed abundant losses
of phosphoric acid and lower sequence coverage.51 A possible
explanation for the minimized loss of phosphoric acid for SPITC-
derivatized peptides in comparison to unmodified peptides upon
(51) When KRpTIRR was guanidinated before N-terminal derivatization, the
already extremely basic peptide became even more basic with the
conversion of the lysine to a homoarginine residue. This significantly
reduced the abundance of the doubly charged ion, making it difficult to
analyze; therefore, KRpTIRR was not guanidinated to ensure comparison
of the ETcaD method with CID. This was the only lysine-containing peptide
not guanidinated in this study.
3650 Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 4. Comparison of ETcaD, CID, and ETD results for multiply charged tryptic peptides from myoglobin obtained by LC-MSn. SPITC-
derivitized peptides are denoted by (SPITC), and unmodified peptides by (unmod.). The peak areas of the product ions for all C-terminal ions
and all N-terminal ions were summed separately and converted to percentages. All lysines are guanidinated for SPITC modified peptides. Φ
denotes peptides with high apparent C-terminal ion selectivity upon CID but which was actually due to production of very dominant yn-1 ions and
not due to production of a comprehensive series of y ions as would be needed for effective de novo sequencing.
ETcaD is that the loss of SPITC or SPTIC + ammonia moieties
are even lower energy pathways in comparison to the loss of
phosphoric acid, rendering the latter pathway less competitive for
the SPITC-derivatized peptides.
One obvious advantage is that the more highly charged, larger
peptides provide greater protein sequence coverage, and their
fragmentation patterns can only be successfully simplified by the
ETcaD technique. Also as previously described, the z ions
produced during ETcaD retain labile PTMs. In contrast, CID of
the even-electron, singly charged SPITC-derivatized peptides
produced fragment ions which did not retain PTMs and offered
lower sequence coverage for phosphopeptides. Furthermore, a
highly basic residue (e.g., arginine or homoarginine) is needed
at the C-terminus to enhance the y ions for CID analysis of SPITC-
modified peptides.14 Without these highly basic residues, the
proton is not sequestered on the C-terminus (i.e., it is more
mobile), resulting in a more diverse array of both C-terminus and
N-terminus ions and lower sequence coverage. However, a highly
basic arginine or homoarginine is not needed at the C-terminus
when using ETcaD due to the different radical dissociation
mechanism inherent to electron-based methods, yet the resulting
The peptide TRDIYETDYpYRK was also N-terminally deriva-
tized with SPITC and subjected to ETcaD analysis. Like the
peptide KRpTIRR, SPITC-TRDIYETDYpYRK showed very high
sequence coverage upon ETcaD with little neutral loss of
phosphoric acid, which is illustrated for the triply charged species
in Figure 3 in the Supporting Information. These results illustrate
that the combination of N-terminal derivatization and ETcaD can
simultaneously simplify spectra and aid identification of PTMs.
Benefits of Enhanced z Ions over Enhanced y Ions.
Although ETcaD of multiply charged N-terminally sulfonated
peptides and CID of singly charged N-terminally sulfonated
peptides are complementary methods for data-dependent LC-MS
analyses, there are significant advantages of the former method.
Analytical Chemistry, Vol. 81, No. 9, May 1, 2009 3651

Figure 5. ETcaD of the phosphorylated peptide KRpTIRR (a) derivatized with SPITC (doubly charged), 25 mV subsequent collision voltage,
q ) 0.20; (b) derivatized with SPITC (triply charged), 30 mV subsequent collision voltage, q ) 0.20; (c) unmodified (doubly charged), 21 mV
subsequent collision voltage, q ) 0.20; and CID of KRpTIRR (d) derivatized with SPITC (singly charged), 45 mV collision voltage, q ) 0.25; (e)
derivatized with SPITC (doubly charged), 14 mV collision voltage, q ) 0.25. Product ions due to cleavage of a portion of the SPITC moiety are
denoted by #, and precursor ions for CID spectra are denoted by *. The unmodified peptide is denoted by M.
spectra still offer high sequence coverage and yield a single
dominant series of product ions. This advantage has already been
seen for the peptide GHHEAELKPL in Figure 4. Therefore, the
combination of ETcaD and N-terminal derivatization could poten-
tially be used for spectral simplification of a variety of peptide
mixtures created from proteins using enzymes that do not
necessarily cleave after lysines and arginines.
Formation of Even Electron versus Radical z Ions. Since
both even electron and radical z product ions are formed by
ETcaD, it was relevant to assess the formation preference of these
two types of z ions. These two types can generally be distinguished
from each other as both the even electron and radical ion peaks
are usually present at each z ion position and differ by one mass
unit. However, the ability to predict the preference for formation
of the radical or even electron z ions would enhance the success
of de novo sequencing algorithms since some amino acids differ
in mass by only 1 amu. To gain insight into the z ion distributions,
area counts of all even electron and radical z ions were tabulated
for each z ion (z1, z2, z3 . . . z14) from 15 different peptides
corresponding to a total of 23 different spectra, which included
different charge states and two different CID q-values and
collision voltages. Both SPITC-modified model peptides and
SPITC-modified peptides from tryptic digests of myoglobin and
cytochrome c were included in this facet of the study. The
percentages of even electron z ions relative to the total amount
of even electron and radical z ions are summarized in Figure
6
for each zn product. As illustrated in Figure 6, there is a notable
increase in the formation of even electron z ions as the length of
the z ion decreases. The identity of specific amino acids adjacent
to the cleaved amide bond had little or no impact on the extent of
even electron z ion formation. However, further investigation with
large data sets may be warranted. There was also no trend in the
3652 Analytical Chemistry, Vol. 81, No. 9, May 1, 2009

Figure 6. Percentage of even electron versus radical z ions as a function of z ion length for 15 different peptide precursor ions and a total of
23 different spectra. n ) the total number of spectra that had observable z ions at each specified sequence position.
formation of one type of z ion when using lower or higher energy
ETcaD conditions.
compared to CID of N-terminally derivatized species but also
compared to electron transfer based methods (i.e., ETD, ETcaD)
of unmodified peptides. The enhanced series of z ions was shown
to retain labile phosphoric acid moieties and also allow efficient
sequencing of a broad range of peptides, including proline-rich
ones and nontryptic species. High sequence coverage was attained
for all peptides with the largest limiting factor being the low mass
cutoff inherent to the ion trap, which sometimes limited the
detection of the lowest m/z ions. With the use of a photodisso-
ciation method, such as IRMPD instead of CID during the
subsequent activation step, it is anticipated that the LMCO
problem could be alleviated.
The proposed mechanism for the formation of even electron
z species from z radical species has been described previously,
and it involves the formation of a hydrogen bond between the z
radical and a c even electron species (either in an intramolecular
or intermolecular process).52 The z radical strips a hydrogen atom
from the c ion, forming an even electron z ion and a radical c
product.52 This pathway is consistent with the trend seen in Figure
6: the longer z ions may have greater opportunity to fold and form
stabilizing intramolecular hydrogen bonds, thus suppressing
hydrogen bond formation with c ions and thus reducing even
electron ion formation. For the case of SPITC-derivatized peptides,
the c products expected upon separation of the z and c species
are neutralized instead of remaining positively charged.
ACKNOWLEDGMENT
Funding from the NSF (Grant CHE-0718320) and the Welch
Foundation (Grant F1155) is gratefully acknowledged.
SUPPORTING INFORMATION AVAILABLE
Additional information as noted in text. This material is
available free of charge via the Internet at http://pubs.acs.org.
CONCLUSIONS
ETcaD of multiply charged, N-terminally sulfonated peptides
results in significant spectral simplification via almost exclusive
formation of z-type ions. This method offers advantages not only
Received for review January 14, 2009. Accepted March 10,
2009.
(52) O’Connor, P. B.; Lin, C.; Cournoyer, J. J.; Pittman, J. L.; Belyayev, M.;
Budnik, B. A. J. Am. Soc. Mass Spectrom. 2006, 17, 576–585
.
AC9000942
Analytical Chemistry, Vol. 81, No. 9, May 1, 2009 3653