High-yielding and controlled dissociation of glycosides producing B- and C-ion species under collision-induced dissociation MS/MS conditions and use in structural determination

Article abstract of DOI:10.1021/ac701686n

Collision-induced dissociation (CID) in mass spectrometry is a powerful technique with which to understand gasphase chemical reactions. A mass spectrometer is used to carry out the reaction, isolation, and analysis. On the other hand, structural analysis

Full text of DOI:10.1021/ac701686n

Anal. Chem. 2007, 79, 9022-9029
High-Yielding and Controlled Dissociation of
Glycosides Producing B- and C-Ion Species under
Collision-Induced Dissociation MS/MS Conditions
and Use in Structural Determination
Katsuhiko Suzuki,† Shusaku Daikoku,† Takuro Ako,† Yuki Shioiri,‡ Ayako Kurimoto,† Atsuko Ohtake,†
Sujit K. Sarkar,† and Osamu Kanie*,†,‡
Mitsubishi Kagaku Institute of Life Sciences (MITILS), 11 Minamiooya, Machida-shi, Tokyo 194-8511 Japan, and
Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, 4259 Nagatsuta-cho,
Midori-ku, Yokohama 226-8501 Japan
thus has potential to form an enormous number of structures.1
Despite the difficulties encountered in structural determination,
it has been reported that glycosylation as a posttranslational
modification (PTM) of protein affects the secretion of glycopro-
tein.2,3 The involvement of PTM in cellular polarity has also been
reported.4,5 It is widely recognized that changes in the composition
of glycosphingolipids are associated with development and dif-
ferentiation.6 Considering the need to analyze minute amounts of
glycoconjugates obtained from biological samples as well as the
existence of glycoforms, it is believed that the use of mass
spectrometry-based analytical methods in the analysis of glyco-
conjugates is a logical consequence.7-9 Structural analysis based
on collision-induced dissociation (CID) is a very powerful tech-
nique in this regard.10,11 It is known that there are spectral
differences among isomeric compounds. In most cases, there are
some differences in signal intensities, while in other cases,
different fragment ions can be found. These phenomena are now
used in a more systematic manner to identify oligosaccharides
by spectral matching.11-17 In structural elucidation, descriptive ion
Collision-induced dissociation (CID) in mass spectrom-
etry is a powerful technique with which to understand gas-
phase chemical reactions. A mass spectrometer is used
to carry out the reaction, isolation, and analysis. On the
other hand, structural analysis of glycan structures is of
extreme importance in the analysis of biomolecules, such
as glycoproteins and glycolipids. In the analysis of glycan
structures based on CID, certain ion species, including
B-/Y-, C-/Z-, and A-/X-ions, are produced. Among these
ions, we are interested in C-ion species that carry a
glycosyl oxygen atom at the anomeric center and that
possibly provide information regarding anomeric config-
uration. A method for generating C-ion species when
necessary is thus considered to be important; however,
none is currently available. In this study, synthetic gly-
cosides carrying a series of aglycons were analyzed with
the aim of identifying suitable glycosides with which to
produce C-ions to be used in the structural determination
of oligosaccharides. The results showed a 4-aminobutyl
group was an excellent candidate. Furthermore, the use
of C-ion species obtained in this manner in the structural
characterization of a ganglioside, GM3, is described. The
type of glycoside is believed to be valuable not only in
structural analysis but also in biological investigation,
because of the existing amino functionality that has been
proven to be useful by enabling the generation of conju-
gates with other molecules and materials.
(1) Laine, R. A.; Pamidimukkala, K. M.; French, A. D.; Hall, R. W.; Abbas, S.
A.; Jain, R. K.; Matta, K. L. J. Am. Chem. Soc. 1988, 110, 6931-6939.
(2) Dube´, S.; Fisher, J. W.; Powell, J. S. J. Biol. Chem. 1988, 263, 17516-
17521.
(3) Roth, J. Chem. Rev. 2002, 102, 285-303.
(4) Rodriguez-Boulan, E.; Gonzalez, A. Trends Cell Biol. 1999, 9, 291-294.
(5) Benting, J. H.; Rietveld, A. G.; Simons, K. J. Cell Biol. 1999, 146, 313-320.
(6) Yamashita, T.; Wada, R.; Sasaki, T.; Deng, C.; Bierfreund, U.; Sandhoff, K.;
Proia, R. L. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 914209147.
(7) Dell, A. Adv. Carbohydr. Chem. Biochem. 1987, 45, 19-72.
(8) Zaia, J. Mass Spectrom. Rev. 2004, 23, 161-227.
(9) Harvey, D. J. Mass Spectrom. Rev. 1999, 18, 349-450.
(10) Hayes, R. N.; Gross, M. L. Collision-Induced Dissociation. In Mass
Spectrometry; McCloskey, J. A., Ed.; Methods in Enzymology 193; Academic
Press: San Diego, CA, 1990; pp 237-263.
(11) Dell, A.; Morris, H. R. Science 2001, 291, 2351-2356.
(12) Viseux, N.; de Hoffmann, E.; Domon, B. Anal. Chem. 1998, 70, 4951-
4959.
(13) Takegawa, Y.; Deguchi, K.; Ito, S.; Yoshioka, S.; Sano, A.; Yoshinari, K.;
Kobayashi, K.; Nakagawa, H.; Monde, K.; Nishimura, S.-I. Anal. Chem.
2004, 76, 7294-7303.
The structural characteristics of oligosaccharides are quite
different from other biopolymers such as nucleic acids and
peptides. Diversity is generated from sequential combination in
the latter two types of polymers, whereas anomers, linkage
position, branching, and sequence are factors in oligosaccharide
structure. Furthermore, the molecular diversity of oligosaccha-
rides is described using the term glycoform. This class of molecule
(14) Kameyama, A.; Kikuchi, N.; Nakaya, S.; Ito, H.; Sato, T.; Shikanai, T.;
Takahashi, Y.; Takahashi, K.; Narimatsu, H. Anal. Chem. 2005, 77, 4719-
4725.
(15) Ashline, D.; Singh, S.; Hanneman, A.; Reinhold, V. N. Anal. Chem. 2005,
77, 6250-6262.
* To whom correspondence should be addressed. Fax: (+81)42-724-6317.
E-mail: kokanee@mitils.jp.
† Mitsubishi Kagaku Institute of Life Sciences (MITILS).
‡ Tokyo Institute of Technology.
9022 Analytical Chemistry, Vol. 79, No. 23, December 1, 2007
10.1021/ac701686n CCC: $37.00 © 2007 American Chemical Society
Published on Web 10/17/2007

Scheme 1. Precursor Ion of a Glycan Producing a Series of Fragment Ions under CID Conditions
species, such as B-/Y-, C-/Z-, and A-/X-ions, are analyzed (Scheme
1).18 The elucidation of a totally unknown structure that is not
present in a database is an important problem that still remains
after introduction of this efficient methodology. In such a case,
an approach based on principal component analysis is considered
to be effective; however, the source of information to be analyzed
has been far from adequate and consists of limited structural
isomers only.19,20 For this reason, a combinatorial oligosaccharide
library should be used as a source of information.21-26 A library
consisting of isomeric compounds with logical structural arrange-
ments should provide indispensable information regarding struc-
tural characteristics. We have previously reported that important
information could be obtained from library compounds based on
energy-resolved mass spectrometry (ERMS).27,28 In order to obtain
fragment ions possessing useful information, we investigated
glycosides carrying various aglycons and determined that some
of the aglycons specifically provide B- and C-ions. Although B-ions
are produced from various glycosides, no report has described
the production of C-ions despite the fact that the ions are often
obtained as a fragment under CID conditions. We report herein
our findings regarding high-yielding and selective dissociation of
sodiated glycosides; that is, 4-bromobutyl and 4-(n-octyloxy)phenyl
glycosides are useful for producing B-ions and 4-aminobutyl
glycoside produces C-ions in good yield.
Analytical thin-layer chromatography was performed on Merck
Art. 5715, Kieselgel 60 F254, 0.25-mm-thickness plates. Visualization
was accomplished with UV light and phosphomolybdic acid or
sulfuric acid solution followed by heating. Column chromatogra-
phy was performed with Merck Art. 7734 silica gel 60, 70-230
mesh. Optical rotations were measured in a 1.0-dm tube with a
Horiba SEPA-200 polarimeter. 1H NMR (500 MHz) and 13C NMR
(125 MHz) spectra were recorded with an Avance 500 spectrom-
eter (Bruker Biospin Inc.) in deuterated solvents using tetram-
ethylsilane as an internal standard.
Synthesis of Compounds 1R/â-8R/â. The detailed procedures
and structural characterizations for the synthesis of compounds
1R/â-8R/â are described in the Supporting Information. For
structures, see Chart 1. A typical procedure is described for the
precursor of compound 5 (12) as follows (Scheme 2). A
phenylthio glycoside of lactose (9) was treated with NaH and
BnBr in N,N-dimethylformamide (DMF) to give perbenzylated
lactoside (10), which was then dissolved in dichloroethane (DCE)
glycosylated with benzyloxycarbony (Cbz) protected 4-aminobu-
tanol in the presence of N-iodosuccinimide (NIS), and a catalytic
amount of trifluoromethanesulfonic acid (TfOH) to afford the
glycosylated compound 11 in moderate yield. Compound 11 was
hydrogenated, using Pd(OH)2 on charcoal in the presence of di-
tert-butyldicarbonate (Boc2O), to provide a mixture of compounds
(12). The Cbz group was first removed to give an amino
functionality, which was then protected in situ with a Boc
group. Compounds 12R and 12â were finally separated using
HPLC.
Derivatization of GM3. A ganglioside, GM3 (>98%, bovine
brain), was purchased from HyTest Ltd. (Turku, Finland). GM3
was esterified using MeI (extra pure; Nacalai Tesque, Inc., Kyoto,
Japan) in dimethyl sulfoxide (extra pure, Merck, Darmstadt,
Germany).
Solvents for Mass Analysis. For analyses using a mass spec-
trometer, HPLC grade solvents were used; MeOH and CHCl3
(Wako Pure Chemical Industries, Ltd, Osaka, Japan). Deuterated
methanol-d4 (NMR Grade) was used for the D-H exchange
experiments.
Instrumentation and Data Collection. Samples of disaccharides
were analyzed using a quadrupole ion trap mass spectrometer
coupled with an electrospray interface (Bruker Esquire 3000 plus,
Bruker Daltonics GmbH, Bremen, Germany). Samples dissolved
in MeOH (0.01-0.1 µmol/mL) were introduced into the ion
source via infusion (flow rate, 120 µL/h). The parameters for the
analysis were as follows: (1) “dry temperature”, 250 °C; (2)
nebulizer gas (N2), 10 psi; (3) dry gas (N2), 4.0 l/min; (4) “Smart
frag.”, off; (5) scan range, m/z 50-750; (6) compound stability,
MATERIALS AND METHODS
Materials. General Method for Chemistry. Dried solvents were
used for all chemical reactions. Solutions were evaporated under
reduced pressure at a bath temperature not exceeding 50 °C.
(16) Zhang, H.; Singh, S.; Reinhold, V. N. Anal. Chem. 2005, 77, 6263-6270.
(17) Lapadula, A. J.; Hatcher, P. J.; Hanneman, A. J.; Ashline, D. J.; Zhang, H.;
Reinhold, V. N. Anal. Chem. 2005, 77, 6271-6279.
(18) Domon, B.; Costello, C. E. Glycoconjugate J. 1988, 5, 397-409.
(19) Fa¨ngmark, I.; Jansson, A.; Nilsson, B. Anal. Chem. 1999, 71, 1105-1110.
(20) Higgins, M. K.; Bly, R. S.; Morgan, S. L. Anal. Chem. 1994, 66, 2656-
2668.
(21) Kanemitsu, T.; Wong, C.-H.; Kanie, O. J. Am. Chem. Soc. 2002, 124, 3591-
3599.
(22) Suzuki, K.; Ohtsuka, I.; Kanemitsu, T.; Ako, T.; Kanie, O. J. Carbohydr. Chem.
2005, 24, 219-236.
(23) Kanie, O.; Ohtsuka, I.; Ako, T.; Daikoku, S.; Kanie, Y.; Kato, R. Angew. Chem.,
Int. Ed. 2006, 45, 3851-3854.
(24) Ohtsuka, I.; Ako, T.; Kato, R.; Daikoku, S.; Koroghi, S.; Kanemitsu, T.; Kanie,
O. Carbohydr. Res. 2006, 341, 1476-1487.
(25) Kanemitsu, T.; Daikoku, S.; Kanie, O. J. Carbohydr. Chem. 2006, 25, 361-
376.
(26) Ako, T.; Daikoku, S.; Ohtsuka, I.; Kato, R.; Kanie, O. Chem. Asian J. 2006,
1, 798-813.
(27) Daikoku, S.; Ako, T.; Kurimoto, A.; Kanie, O. J. Mass Spectrom. 2007, 42,
714-723.
(28) Daikoku, S.; Ako, T.; Ohtsuka, I.; Kato, R.; Kanie, O. J. Am. Soc. Mass
Spectrom. 2007, 18, 1876-1882.
Analytical Chemistry, Vol. 79, No. 23, December 1, 2007 9023

Chart 1
experiment); the first and the last two data sets, which are
associated with a transient period to steady state, in an rf amplitude
step were not used in order to avoid any inaccuracy.
Isotopic peaks with [Ii + 1] and [Ii + 2], where In indicates a
fragment ion, were included in the calculations. (See also Data
Handling.) For the isolation of a product ion, m/z (2 (w ) 2)
were isolated and subjected to the CID experiments to include
isotopes. Standard MS/MS spectra are the extracts of these ERMS
at a designated amplitude.
Data Handling. In order to obtain graphs of the ERMS, the
following equations were used. When an ion “IP” produces a series
of product ions, I1, I2, I3,... Ii, the relative ion currents for individual
ions were defined by the equation,
i
CI
relC )
× 100
(1)
(2)
P
CI
i
CI
relC )
× 100
i
n
P
CI + CI
∑
i)1
where relC indicates the ion current of a given ion among observed
i
ions in percentage, CI is the observed ion current in focus, and
P
CI is the ion current of a precursor ion. The calculations were
performed using a program we developed with Excel (Excel 2000
(Microsoft Co.)), which was based on DSUM function and
programmed to choose a range of isotopes (w) to be taken into
consideration (w ) 2 in the experiments).
RESULTS AND DISCUSSION
As explained above, B- and C-ion species are often observed
under CID conditions during structural investigation of oligosac-
charides. For this reason, if these ion species of specific sequences
can be made readily available, they can be used as references for
comparisons in the structural determination. We have been
attempting to identify suitable aglycons to be used in structural
elucidation, providing such glycon-oriented ion species when
needed. ERMS is a very powerful tool for the structural analy-
sis of isomeric compounds such as oligosaccharides and is
considered to be a suitable method with which to examine the
details of the dissociation profile under CID conditions.29-40 With
Scheme 2. Synthesis of Compound 12
(29) McLuckey, S. A.; Glish, G. L.; Cooks, R. G. Int. J. Mass Spectrom. Ion Phys.
1981, 39, 219-230.
(30) Fetterolf, D. D.; Yost, R. A. Int. J. Mass Spectrom. Ion Phys. 1982, 44, 37-
50.
(31) Verma, S.; Ciupek, J. D.; Cooks, R. G. Int. J. Mass Spectrom. Ion Proc. 1984,
62, 219-225.
(32) Bursey, M. M.; Nystrom, J. A.; Hass, J. R. Anal. Chim. Acta 1984, 159,
265-274.
300%; (7) ICC target, 5000; (8) maximum acquisition time,
200 ms; (9) average, 10 spectra; and (10) “cutoff”, 27.6% of the
corresponding precursor ion.
In our MSn experiments, the end cap rf amplitude was raised
by 0.02-V increments until the precursor ion could no longer be
detected (plateau at less than 0.9% of total ion current). Only end
cap rf amplitude was controlled during the CID experiments. The
He pressure was 4.86 × 10-6 mbar, and the CID time was 40 ms.
Averages of m - 4 spectra were used for CID experiments (m )
14: where m is the number of spectra obtained during the
(33) Favretto, D.; Guidugli, F.; Seraglia, R.; Traldi, P.; Ursini, F.; Sevanian, A.
Rapid Commun. Mass Spectrom. 1991, 5, 240-244.
(34) Williams, J. D.; Cox, K. A.; Cooks, R. G.; McLuckey, S. A.; Hart, K. J.;
Goeringer, D. E. Anal. Chem. 1994, 66, 725-729.
(35) Hart, K. J.; McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1994, 5, 250-259.
(36) Gabelica, V.; Galic, N.; De Pauw, E. J. Am. Soc. Mass Spectrom. 2002, 13,
946-953.
(37) Crowe, M. C.; Brodbelt, J. S. J. Am. Soc. Mass Spectrom. 2003, 14, 1148-
1157.
(38) Broeren, M. A. C.; van Dongen, J. L. J.; Pittelkow, M.; Christensen, J. B.;
van Genderen, M. H. P.; Meijer, E. W. Angew. Chem., Int. Ed. 2004, 43,
3557-3562.
9024 Analytical Chemistry, Vol. 79, No. 23, December 1, 2007

Figure 1. Representative energy-resolved mass spectra for [M + Na]+. Representative fragment ions: red, TIC; green, precursor ion; blue,
B2-ion (m/z 347); beige, Y1-ion; pink, butenyl compound; light green, C2-ion (m/z 365).
this aim in mind, we prepared a series of compounds (1R/â-
8R/â). (Details of syntheses are provided in Supporting Informa-
tion (SI) section 1).
one-third of the B2-ions, whereas its counterpart, â-glycoside (1â)
produced Y1-ion as a major fragment (Figure 1A,B). This type of
difference in fragmentation depending on anomeric configuration
is often observed and is useful in discriminating anomers.1,27,28,39,41
Here, we have focused on the overall ionic profiles where the total
ion current (TIC) gradually decreased to 10.3 and 4.0% for octyl
R- and â-lactosides, respectively, over a range of 0.7 V during the
CID experiments (This amplitude range was determined based
on the observation that all compounds used in the experiments
dissociated.). This drastic decrease in “ion quantity” will most
likely be problematic in future glycan analysis using an oligosac-
charide library as a structural information source. Therefore, the
ion recovery needs to be improved. Furthermore, B2- and C2-ions
in these cases will be used for further MS/MS analysis to provide
more detailed information regarding the structures of these ions.
These data obtained at the MSn stage of CID experiments are
Analysis of Energy-Resolved Mass Spectrum of a Series
of Glycosides. We conducted CID experiments using these
compounds to determine suitable glycosides that produce B- and
C-ion species using a disaccharide (lactose) structure as an
example. Sodiated precursor ions ([M + Na]+) were collided with
He gas in an ion trap mass spectrometer to produce a series of
product ions with various end-cap electrode amplitudes. Repre-
sentative ERMS thus obtained are shown in Figure 1, and the
fragmentation pathways during the CID process are depicted in
Scheme 3. An n-octyl R-lactoside (1R) produced B1-ion (m/z 347)
as a major fragment ion followed by Y1-ion (m/z 315) in about
(39) Kurimoto, A.; Daikoku, S.; Mutsuga, S.; Kanie, O. Anal. Chem. 2006, 78,
3461-3466.
(40) Kurimoto, A.; Kanie, O. Rapid Commun. Mass Spectrom. 2007, 21, 2770-
2778.
(41) Yamagaki, T.; Fukui, K.; Tachibana, K. Anal. Chem. 2006, 78, 1015-1022.
Analytical Chemistry, Vol. 79, No. 23, December 1, 2007 9025

Scheme 3. Generalized Fragmentation
Pathways of the CID Process of Lactosyl
Glycosides
Table 1
generated ion species, %
B2-ion
Y1-ion
0,3A2-ion
C2-ion
(m/z 347)
(Gal-cleavage)
(m/z 305)
(m/z 365)
compd
R
â
R
â
A
â
R
â
1
2
3
4
5
6
7
8
9.4
0.6
3.5
0.0a
11.0
11.4
5.9
4.2
0.0a
11.2
10.9
6.9
10.7
12.0
2.8
0.4
0.7
0.0a
0.0a
0.0a
0.0a
26.7
4.3
0.0a
0.0a
0.0a
0.2
99.5 79.2
0.0a 0.0a
16.2
16.4
15.8
21.0
17.9
3.6
7.5
7.8
3.7
2.8
0.4
0.5
3.8
2.1
2.4
3.0
23.3
3.5
9.4
1.0a 1.5
0.0a 1.0
0.0a 0.0a
8.6
1.6
0.4
0.2
50.0 12.4
0.0a
0.0a
a Not observed. Yields were calculated at an amplitude range of 0.1
V where the graphs of a series of ions were maximum.
nism proceeding via a possible five-membered transition-state. The
next examples are 4-aminobutyl glycosides (5R and 5â) obtained
from the Boc-protected precursor at MS2, where we expected a
long-distance hydrogen abstraction reaction might take place
through a cyclic transition state (Scheme 4 B, path a). (See SI,
section 2.1 for the generation of the ion from the precursor ion
5.) In this reaction, we anticipated that a pyrollidine would be
released when a C2-ion species is generated. The ERMS indicated
that the C2-ion (sodiated lactose equivalent) was indeed a main
product in both anomers (Figure 1E,F). The ion yields obtained
for aminobutyl R- and â-glycosides were 22.6 and 18.1%, respec-
tively, over a 0.7-V range. Although the ion yields were not
comparable to those of bromobutyl glycosides over this range of
rf amplitudes, MS/MS analysis is usually carried out at an rf
amplitude where the desired ion species is produced in the
greatest amount, which is ∼1.2 V in each anomer in this case.
The ion yields at 1.2 V were 69.9 and 59.3%. A series of fragment
ions was obtained where C2-ion was the most abundant.
Possible Dissociation Mechanism. In order to ascertain the
mechanism of this dissociation reaction, the effect of chain length
on the dissociation reaction was evaluated based on ERMS. It was
found that the generation profiles of C2-ions from sodiated
compounds 3-7 were similar to those in the radical cyclization
reactions (Table 1). Furthermore, the ratio of the C2-ion yields
from R- and â-lactosides is 1.1, which indicates that dissociation
of the aglycon was independent of the stereochemistry at C-1.
This is in contrast to the fact that the energy requirement for the
R- and â-glycosides differs27,28,39 and supports the reaction mech-
anism shown in Scheme 4B. Since no radical species were
observed, it is believed that the reaction proceeded in a concerted
fashion or the hydrogen abstraction process was completed within
an experimental period (40 ms). We looked for a pyrrolidine,
possibly produced in the reaction, since it would represent strong
evidence; however, one was not observed. However, a CID
experiment using a deuterated compound did produce a deuter-
ated C2-ion, indicating that deuterium was transferred from an
amino group and strongly supporting our model (SI; section 2.2).
In the dissociation, a considerable amount of A-series ions was
observed. A-Ion species are usually generated from hemiacetals;
therefore, we believe that the observed A-ions were produced via
C2-ion (Scheme 3). Regardless of the actual mechanism, it should
be noted that this is the first report that has described obtaining
C-ions based on a mechanism-based dissociation reaction.
useful in structural analysis based on “spectral matching” between
a fragment ion generated from a glycan and one from a library.
Dissociation of 4-Bromobutyl Lactoside. We examined the
CID of Na+ adducts of 1-bromobutyl R- and â-lactosides (2R and
2â). The ERMS are presented in Figure 1C,D. As is clear from
the spectra, the “ion yields” for both anomers were improved,
which was surprisingly found to be quasi-quantitative over a 0.7-V
range. It was found that B2-ion is generated preferentially in the
dissociation of bromobutyl glycosides. It is thought that lower
activation energy was required for the cleavage of a Br-C bond
and is responsible for the high-yield reaction. We believe that the
difference in dissociation energy for Br-C bonds and coordination
of Na+ on multiple oxygen atoms is particularly important because
desired ion species can be generated before Na+ dissociates
leaving a neutral molecule. Also, involvement of a cyclic interme-
diate was suggested in this preferable dissociation because a
sodiated lactosyl bromide (m/z 427) was observed as well in an
R-glycoside (Scheme 4A). The presence of a sodiated butenyl
glycoside (m/z 419) observed in both anomers suggested the
involvement of three competing paths, B-Ion generation (path a),
bromination (path b), and an elimination (path c), in the dissocia-
tion. Regardless of the transition states and stereochemistry of
the products, it was shown that the bromobutyl glycoside might
be useful in structural analysis by providing B-ion.
Dissociation of 4-Aminobutyl Lactoside. We further inves-
tigated the dissociation of glycosides based on a similar mecha-
9026 Analytical Chemistry, Vol. 79, No. 23, December 1, 2007

Scheme 4. Anticipated Reaction Mechanism of 4-Substituted Butyl Lactosides.
Further analysis of ERMS showed that the profiles of B2-ion
generation are similar among n-octyl (1), 4-aminobutyl (5), and
octyloxyphenyl glycosides (8), with the exception of 4-bromobutyl
glycoside (2) (Figure 2). The general tendency that R-glycosides
in each set required less activation energy (provided by rf
amplitude) is in agreement with our previous results, which
indicated a dependency on anomeric configuration. The profile
of B2-ion generation for 2 (Figure 2B) therefore suggests that
the mechanism of the bromobutyl group occurred independently
of the anomeric configurations (Scheme 4A). The generation
profile for C2-ion in 5 (Figure 2D) is similar to that of 2, which
indicates again that the dissociation reaction to produce the ion
species is independent of anomeric configurations and supports
our proposed mechanism (Scheme 4B).
One of the possible reasons that neither a furan nor a
pyrrolidine was detected in these reactions was that the reactions
were not promoted by the Na+, as is usually the case. In this
model, Na+ is coordinated most likely at multiple oxygen atoms
between Gal and Glc, but the dissociation took place at a location
remote from the site around the aglycon. We conducted more
extensive synthetic and CID experiments and confirmed the
formation of pyrrolidine (Data will be reported elsewhere.).
Contrary to the dissociations of sodiated compounds, no C2-ion
was observed in the dissociation of protonated 5â (Scheme 4C).
Figure 2. ERMS of sodiated compounds 1, 2, 5, and 8. Ratios of individual product ions (m/z 347 and 365) over respective precursors were
calculated to enable discussion of generation profiles. Closed circle, R-glycoside; open circle, â-glycoside. The relative ion current (relC) was
calculated based on eq 1.
Analytical Chemistry, Vol. 79, No. 23, December 1, 2007 9027

Both Y-ions that originated with a reducing terminus were
observed, although no nonreducing end associated ion (B- and
C-ions) was observed. (SI; section 2.3.1) This suggests an amino
group is protonated, and yet glycosidic bond cleavages took place.
The above two reactions can be explained by a charge remote
process. This might be related to our and other observations that
R-glycosides are labile under CID conditions, which is the opposite
of commonly found protonation-associated solution chemis-
try.27,28,39,41 Similar phenomena can be found for a pair of pyridy-
laminated oligosaccharides as well (SI; section 2.3.2). Although
other mechanisms including cation-induced dissociation may exist,
the possibility of a charge-independent dissociation mechanism
should be considered.
With regard to bromobutyl glycosides, their high potential
value for producing B-type ion species in good ion yield is going
to be useful in the future development of MS-based analytical
methods for glycan structures. It was also found that a phenyl
glycoside (8), especially R-glycoside (8R), produced B-ions in a
good yield (50.0%) (Table 1) and that it could be an alternative to
a bromobutyl glycoside when a synthesis does not permit its use
(Figure 1G,H).
Figure 3. ERMS of C-ion species from [5R + Na]+ and [5â + Na]+.
(A, B) ERMS spectra obtained for C2-ions from 4-aminobutyl R- and
â-lactosides, respectively. The relative ion current (relC) was calculated
based on eq 2.
C-Ions from a Pair of Anomers. We examined a series of
ERMS spectra obtained for R- and â-aminoalkyl glycosides of
Figure 4. MS/MS analyses of a ganglioside GM3. (A) MS/MS analysis of sodiated GM3 methyl ester; (B) MS/MS of Y2-ion (sodiated lactosyl
ceramide).
9028 Analytical Chemistry, Vol. 79, No. 23, December 1, 2007

shows that the C2-ion (m/z 365) was produced while the abundant
ion was the Y1-ion (m/z 751). C2-Ion obtained in this manner was
then analyzed by means of ERMS, as shown in Figure 5.
Comparison of the ERMS and ones obtained from R- and
â-aminobutyl glycosides of lactose (Figure 3) resulted in confirma-
tion that the C2-ion derived from the ganglioside GM3 had indeed
a â-glycosidic linkage between glucose and the ceramides. This
suggests that C-ion species are useful in the structural determi-
nation of fragments obtained at MSn and also indicates the
importance of an aglycon that generates C-ions.
Figure 5. ERMS of C2-ion obtained after MS3 of GM3. The relative
ion current (relC) was calculated based on eq 2.
CONCLUSION
We succeeded in identifying very useful aglycons that provide
a series of fragment ions that are useful in MS/MS-based
structural analysis. It has been impossible until now to obtain C-ion
species when needed. A 4-aminobutyl glycoside was introduced
for the first time to overcome this problem. The use of a specific
type of aglycon is advantageous because of their adaptability in
solid-phase extraction, when they are suitably protected, and their
potential to couple with other materials using the amino group
after deprotection. Thus, it is hoped that the use of the type of
glycoside will not be limited to analyses but will also be very useful
when constructing conjugates and arrays for functional investiga-
tion. The importance of C-ion species in the structural determi-
nation of naturally occurring glycoconjugates was also presented
by confirming the â-lacotsyl linkage in the GM3.
lactose focusing on the generation of C2-ions and found that the
4-aminobutyl glycosides are excellent candidates for this particular
purpose. We now discuss the stereochemistry of thus obtained
C2-ion species, the chemical structures of which resemble sodiated
R- and â-hemiacetals, and in this case lactose. The individual C2-
ions were further examined under CID conditions (ERMS), and
the analyses resulted in completely different spectra. (Figure 3A,B)
According to the fragmentation process of the respective precur-
sors and the fragment ions generated (see fragments presented
in Scheme 3), we are convinced that both C-ions are lactoses,
although with different anomeric configurations. This indicates
that anomers of hemiacetals have distinct behaviors and are
distinguishable from each other, and furthermore, this suggests
the C-ion species can be used in the structural determination of
fragments, including their anomeric configurations. In order to
utilize this finding, we decided to perform a CID analysis of a
ganglioside, GM3, which contains a â-lactosyl glycoside in its
structure. We conducted a CID experiment of a sodiated ion of a
methyl ester of GM3 (m/z 1218)42 from bovine brain that produced
Y2-ion (m/z 913), the structure of which resembled a sodiated
lactosyl ceramide, as the most abundant ion. (Figure 4A) We used
a methyl ester of GM3 because it is known that the glycosidic
bond of esterified sialic acid becomes more stable under CID
conditions, which would enable the production of various ion
species. Among the fragment ions generated, the Y2-ion was
isolated and further analyzed under CID conditions. Figure 4B
ACKNOWLEDGMENT
Financial supports from Mitsubishi Chemical Co. (MCC) and
the Key Technology Research Promotion Program of the New
Energy and Industrial Development Organization (NEDO) of the
Ministry of Economy, Trade and Industry of Japan is acknowl-
edged.
SUPPORTING INFORMATION AVAILABLE
Additional information as noted in text. This material is
available free of charge via the Internet at http://pubs.acs.org.
Received for review August 8, 2007. Accepted September
6, 2007.
(42) Handa, S.; Nakamura, K. J. Biochem. 1984, 95, 1323-1329.
AC701686N
Analytical Chemistry, Vol. 79, No. 23, December 1, 2007 9029