Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry analysis of oligosaccharides and oligosaccharide alditols obtained by hydrolysis of agaroses and carrageenans, two important types of red seaweed polysaccharides

Article abstract of DOI:10.1016/j.carres.2009.10.009

MALDI-TOF mass spectrometry analyses of several oligosaccharides (aldoses) and oligosaccharide alditols derived from agaroses, kappa- and iota-carrageenans using different matrices (2,5-dihydroxybenzoic acid, nor-harmane, ferulic acid, and the ionic liqui

Full text of DOI:10.1016/j.carres.2009.10.009

Carbohydrate Research 345 (2010) 275–283
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF)
mass spectrometry analysis of oligosaccharides and oligosaccharide
alditols obtained by hydrolysis of agaroses and carrageenans,
two important types of red seaweed polysaccharides
a
b
c
c
c
_
M. Kaniz Fatema , Hiroshi Nonami , Diogo R. B. Ducatti , Alan G. Gonçalves , M. Eugenia R. Duarte ,
Miguel D. Noseda c, Alberto S. Cerezo d, Rosa Erra-Balsells d, , María C. Matulewicz
d,
*
*
a The United Graduate School of Agricultural Sciences, Faculty of Agriculture, Ehime University, 3-5-7-Tarumi, Matsuyama 790-8566, Japan
b Plant Biophysics./Biochemistry Research Laboratory, College of Agriculture, Ehime University, 3-5-7 Tarumi, Matsuyama 790-8566, Japan
c Departamento de Bioquímica e Biologia Molecular, Universidade Federal do Paraná, PO Box 19046, Curitiba, Paraná, Brazil
d Departamento de Quimica Organica—CIHIDECAR (CONICET-UBA), Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires, Pabellón 2-Ciudad Universitaria,
1428EGA Buenos Aires, Argentina
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 12 June 2009
Received in revised form 6 October 2009
Accepted 13 October 2009
Available online 11 December 2009
MALDI-TOF mass spectrometry analyses of several oligosaccharides (aldoses) and oligosaccharide alditols
derived from agaroses, kappa- and iota-carrageenans using different matrices (2,5-dihydroxybenzoic
acid, nor-harmane, ferulic acid, and the ionic liquid matrices 2,5-dihydroxybenzoic acid–n-butylamine
and ferulic acid–n-butylamine) were conducted. These carbohydrates were selected as model compounds
to study the MALDI prompt and post-source decay (PSD) fragmentation processes of both families of oli-
gosaccharides. Sulfated alditols showed in the negative-ion mode the molecular ion as [MꢀNa]ꢀ together
with the species yielded by their prompt fragmentation (mainly desulfation) while the sulfated oligosac-
charides (aldoses) showed mainly glycosidic prompt fragmentation (glycosidic C-cleavages and desulf-
ation). Non-sulfated aldoses and alditols, which could only be analyzed in positive-ion mode
([M+Na]+), did not suffer any prompt fragmentation. The former yielded cross-ring fragmentation in
the PSD mode. Best results were obtained by using 2,5-dihydroxybenzoic acid and/or nor-harmane as
matrices for all the compounds studied.
Keywords:
Agaro-oligosaccharides MALDI MS
Agaro-oligosaccharide alditols MALDI MS
Kappa-carrageenan oligosaccharides MALDI
MS
Iota-carrageenan oligosaccharide alditols
MALDI MS
Nor-harmane
DHB
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
(3-linked b-
lactopyranose), 60-sulfated agarose (3-linked b-
6-sulfate and 4-linked 3,6-anhydro- -galactopyranose) and
methylated agarose (3-linked 6-O-methyl-b- -galactopyranose
and 4-linked 3,6-anhydro-
-galactopyranose).2–9 The use of par-
D
-galactopyranose and 4-linked 3,6-anhydro-
a-L-ga-
D-galactopyranose
Galactans biosynthesized by red seaweeds (Rhodophyta) essen-
tially consist of linear chains of alternating 3-linked b-
pyranosyl and 4-linked -galactopyranosyl units and, in many
a-L
D
-galacto-
D
a
a-L
cases, the latter residue appears in the 3,6-anhydro form. These
galactans are classified either as carrageenans if the 4-linked
tial hydrolysis methods on repetitive agarans and carrageenans is
an attractive way to obtain oligosaccharides with a specific sulfa-
tion pattern in relative high yields.10–13 These oligosaccharides
can be utilized as useful standards and model compounds for the
study of complex native algal galactans,14 and are also potential
sources of different positional isomers and/or diastereoisomers of
sulfated oligosaccharides which have proved to exhibit a variety
of physiological activities.14,15 The extent of these activities has
been correlated with the degree of polymerization and the percen-
tage and position of the sulfate groups.14,15 Therefore, analytical
techniques capable of resolving complex mixtures of isomeric
oligosaccharides obtained from agarans and carrageenans were
developed and structural elucidation of these oligosaccharides
was mainly carried out by NMR spectroscopy.13,14
residue is in the
L
D-configuration or agarans if this unit is in the
-configuration.1 Most of these polysaccharides are found as anio-
nic polymers with varying degrees of sulfation. Furthermore, some
algal species produce these galactans with an almost idealized
repeating unit such as kappa-carrageenan (3-linked b-
pyranose 4-sulfate and 4-linked 3,6-anhydro- -galactopyra-
nose), iota-carrageenan (3-linked b- -galactopyranose 4-sulfate
and 4-linked 3,6-anhydro- -galactopyranose 2-sulfate), agarose
D-galacto-
a
-D
D
a
-D
* Corresponding authors. Tel./fax: +54 11 45763346.
E-mail addresses: erra@qo.fcen.uba.ar (R. Erra-Balsells), cristina@qo.fcen.uba.ar
(M.C. Matulewicz).
0008-6215/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2009.10.009

276
M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
Recently, MALDI mass spectrometry (MALDI MS) has become a
for this compound in the positive-ion linear and reflectron modes
when nor-harmane was used as matrix.
major technique for the analysis of carbohydrates being their struc-
tural complexity a major problem for these studies.16–18 Particularly,
sulfated oligosaccharides are difficult to analyze by this technique,
partly because of the labile nature of the sulfate group.16–18 Results
can be erratic, indicating that there are still several unknown struc-
tural and experimental factors that affect the measurements.
As part of a research project related with the study of MALDI MS
analysis of carbohydrates,19–24 we report here the behavior of se-
lected model compounds, oligosaccharides (aldoses) and oligosac-
charide alditols, using different matrices [2,5-dihydroxybenzoic
acid (DHB), nor-harmane, ferulic acid (FA), and the ionic liquid
matrices DHB–n-butylamine (DHB–Bu) and FA–Bu]. These com-
pounds, obtained from agaroses, kappa- and iota-carrageenans,
were chosen to study the MALDI and the fragmentation processes
(prompt fragmentation and PSD) of both families of carbohydrates.
Compound 3 gave the molecular ion peak [M+Na]+ at m/z 652.9
only by using DHB as matrix. In the PSD analysis from this precur-
sor ion a quite efficient decomposition process yielded a signal at
m/z 289.8, a protonated species formed after glycosidic B2- and/or
Z2-fragmentation with further loss of H2O (Table 3). The presence
of the G?LA repeating unit in the structure of compound 3 would
account for this behavior. No signals of this compound were
observed in the MALDI mass spectrum by using nor-harmane as
matrix.
Although analyses in the negative-ion mode were attempted by
using DHB and nor-harmane as matrices, compounds 1–3 did not
yield [MꢀH]ꢀ or any stable anion fragment.
The sulfated compounds 4 and 5 (Table 1, Fig. 1) showed a quite
different behavior by MALDI MS analysis. Even though in positive-
ion mode several matrices were assayed, that is, DHB, nor-har-
mane, FA, and the ionic liquid matrices DHB–Bu and FA–Bu, no sig-
nal could be obtained. Results were different in the negative-ion
mode. When DHB was used as matrix, the intact molecular ion
[MꢀNa]ꢀ was only detected for compound 4 at m/z 403.1; with
nor-harmane, together with this signal, peaks at m/z 373.1 due to
the loss of HCHO from the molecular ion and at m/z 258.7 due to
a C1-cleavage, were found with low relative intensity. PSD analysis
of compound 4 using DHB as matrix showed as only fragment the
ion at m/z 241.8 due to a [C1-ionꢀH2O] and/or B1-cleavage of the
glycosidic bond. This peak together with a weak signal at m/z
259.1 attributed to C1-cleavage were observed when nor-harmane
was the MALDI matrix (Table 3). The electrospray-ionization (ESI)
mass spectrum in negative-ion mode of compound 4 gave signals
at m/z 421.2 [M+H2OꢀNa]ꢀ, 403.2 [MꢀNa]ꢀ, 259.1 and 241.2 (both
yielded by prompt fragmentation).
2. Results and discussion
2.1. Oligosaccharides
To investigate the effect of the sulfate group on the stability of
the molecular ions yielded by MALDI, the mass spectra of the neu-
tral oligosaccharides G?LA (1), G?LA?G (2) and G?LA?G?LA
(3) were compared with those of monosulfated G4S?DA (4) and
disulfated G4S?DA?G4S?DA (5) (nomenclature of Knutsen
et al.).3 Experiments were conducted in positive- and negative-
ion linear and reflectron modes and when the intact molecular
ion was detected, in PSD mode too. Table 1 lists the molecular
weights and Figure 1 shows the molecular structures of the studied
oligosaccharides and the main fragmentations (nomenclature of
Domon and Costello is used).25
Finally, disulfated compound 5 did not show the intact molecu-
lar ion as [M+Na]+ and/or [MꢀNa]ꢀ by using DHB and nor-harmane
as matrix (Table 2). In order to overcome this problem, FA and the
ionic liquid matrices DHB–Bu and FA–Bu were assayed. In spite of
the effort, only signals at m/z 710.2, 565.7, 403.2, 258.6 and 241.0
were detected with DHB as matrix. These signals were also ob-
tained, with higher relative intensity, with nor-harmane as matrix
(Fig. 2, Table 2). The signal at m/z 403.4 can be assigned to the
[(G4S?DA)ꢀNa]ꢀ fragment obtained by C2- and/or Y2-fragmenta-
tion of the central glycosidic bond. Fragments observed at m/z
709.5 and 565.0 would be the final species produced after desulf-
ation of the species at m/z 811.6 ([MꢀNa]ꢀ) and the fragment
produced by a C3-type fragmentation respectively. Fragments at
m/z 372.9 and 258.5 also derive from C-fragmentations, but forma-
tion of the former would include shedding of HCHO; the signal at
m/z 240.7could be due to a [C1-ionꢀH2O] and/or B1-cleavage of the
glycosidic bond (Table 2). The presence of these fragments and
those detected by ESI MS of compound 5 (see Section 4) confirms
that the molecular ions produced by both MALDI and ESI are quite
unstable and they suffer prompt fragmentation in the ionization
chamber. In agreement with the behavior observed in MALDI
experiments, ESI analysis only yielded ions in negative-ion mode.
But, on the contrary, by ESI process the stable intact molecular
ion was detected as the double charged species [Mꢀ2Na]2ꢀ at
m/z 394.0. As it is known double charged ions are not detected in
general by MALDI MS analysis. Together with this double charged
anion, an important signal at m/z 403.0 and a small signal at m/z
241.1 were observed. Although MALDI and ESI are soft ionization
methods the minor vibration energy remaining in the gas molecu-
lar ion of 5 is enough to induce its prompt decomposition.
The results obtained in this section of the present study support
our previous suggestion that prompt fragmentation observed in
the MALDI MS analysis of a mixture of kappa-carrageenan oligo-
saccharides was mainly due to glycosidic C-cleavages.20
By using DHB as matrix, compound 1 showed in positive linear
and reflectron modes the intact molecular ion as [M+Na]+ (Table 2)
together with minor signals corresponding to [M+H]+ and the ana-
lyte cluster ion [2M+Na]+. PSD analysis gave as main fragment the
sodiated species at m/z 257.8 produced by a cross-ring 0,3X1 cleav-
age of the galactose ring with shedding of the C3H6O3 (Table 3). It
was impossible to determine spectra with nor-harmane as matrix
since the molecular ion peak [M+Na]+ was overlapped by a matrix
signal at m/z 338.8 with a shoulder at m/z 350.9.
Compound 2 showed in similar experiments the ion peak
[M+Na]+ together with a signal corresponding to the [M+K]+ spe-
cies, being the former the most intense ion. In the PSD mode not
only the fragment produced by cross-ring 0,4X2 and/or 0,4X0 cleav-
age was observed at m/z 448.2, but also the signal at m/z 346.9
due to a C2- and/or Y2-fragmentation. Similar results were obtained
Table 1
List of the oligosaccharides and oligosaccharide alditols analyzed in this study
Compound
Structurea
Molecular weight
Oligosaccharide aldoses
1
2
3
4
5
G?LA
G?LA?G
G?LA?G?LA
G4S?DA
G4S?DA?G4S?DA
324.3
486.4
630.5
426.3b
834.6c
Oligosaccharide alditols
6
7
G?LAOH
G6M?LAOH
326.3
340.3
8
9
10
G6S?LAOH
G4S?DA2SOH
G4S?DA2S?G4S?DA2SOH
428.3
530.4c
1040.7d
Nomenclature of Knutsen et al.3
Monosodiated.
Disodiated.
a
b
c
d
Tetrasodiated.

M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
277
Figure 1. Molecular structure and main fragmentations of oligosaccharides 1–5.
2.2. Oligosaccharide alditols
positive-ion mode as [M+Na]+ (m/z 451.4) and in negative-ion
mode as [MꢀNa] (m/z 405.2); the signal at m/z 286.4, in positive-
ion mode, was attributed to a C- and/or B-fragmentation. PSD anal-
ysis with DHB in negative-ion mode gave a fragment at m/z 242.3
due to a C- and/or B-glycosidic linkage cleavage (Table 3). The anal-
ysis carried out with nor-harmane as matrix, showed similar re-
sults to those described above for DHB. In all cases the spectra
were obtained with quite high S/N ratio.
In spite of the fact that the monosulfated alditol 8 did not show
desulfated ions, even in PSD experiments, the disulfated alditol 9
(MW 530.4, Table 1) gave, by using nor-harmane as matrix, the
molecular ion peak [MꢀNa]ꢀ at m/z 507.8, together with signals at
m/z 405.5, 248.7 and 226.5 (Table 4). The first signal could be ratio-
nalized as due to [MꢀNaSO3+HꢀNa]ꢀ, while the other two could be
produced by a Z-cleavage (Fig. 3). Desulfated fragments were also
observed by PSD analysis (Table 3) at m/z 405.4 and 387.1 (dehydra-
tion of the fragment of m/z 405.4). The spectra obtained by using
DHB as matrix in the negative-ion mode, showed in linear and reflec-
tron modes the above-mentioned signals at m/z 507.6 and 405.6
The behavior of oligosaccharide alditols in MALDI MS analysis
was carefully studied using as model compounds the family listed
in Table 1, constituted by G?LAOH (6), G6M?LAOH (7),
G6S?LAOH (8), G4S?DA2SOH (9), and G4S?DA2S?G4S?DA2-
SOH (10) (Fig. 3).
Compound 6 gave in the positive-ion mode, with DHB and nor-
harmane as matrices, only the peak corresponding to the sodiated
molecular ion [M+Na]+ with m/z 349.5. Spectra obtained with the
DHB showed in general S/N ratio higher than those with nor-har-
mane (Table 4). Compound 6 showed to be quite stable because
no significant fragmentation was observed in PSD experiments.
The analysis of the methylated model compound 7 yielded quite
similar results to those obtained in the former case (Table 4).
The behavior changed drastically when a sulfate group was
present in the alditol structure. Thus, the monosulfated derivative
8 could be analyzed in both positive- and negative-ion modes (Ta-
ble 4). By using DHB as matrix, the molecular ion was observed in

278
M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
Table 2
Molecular and daughter ions observed in the MALDI-TOF mass spectra of oligosaccharides 1–5
Compound
Matrix
DHB
Ion-mode
Positive
m/z (RI)a
Assignmentb
1
347.7 (100%)
325.5 (12%)
524.5 (55%)
508.0 (100%)
[M+Na]+
[M+H]+
[M+K]+
2
DHB
Positive
[M+Na]+
3
4
DHB
Positive
652.9 (100%)
[M+Na]+
Nor-harmane
Negative
403.1 (100%)
373.1 (24%)
258.7 (13%)
403.1 (100%)
[MꢀNa]ꢀ
[MꢀHCHOꢀNa]ꢀ
[C1-ionꢀNa]ꢀ
[MꢀNa]ꢀ
DHB
Negative
Negative
5
Nor-harmane
709.5 (3%)
[MꢀNaSO3+HꢀNa]ꢀ
565.0 (18%)
403.4 (31%)
372.9 (13%)
258.5 (100%)
240.7 (37%)
[C3-ionꢀNaSO3+HꢀNa]ꢀ
[C2-ion and/or Y2-ionꢀNa]ꢀ
[C2-ionꢀHCHOꢀNa]ꢀ
[C1-ionꢀNa]ꢀ
[C1-ionꢀNaꢀH2O]ꢀ and/or [B1-ionꢀNa]ꢀ
DHB
Negative
710.2 (3%)
565.7 (7%)
403.2 (22%)
258.6 (42%)
241.0 (100%)
[MꢀNaSO3+HꢀNa]ꢀ
[C3-ionꢀNaSO3+HꢀNa]ꢀ
[C2-ionꢀNa]ꢀ
[C1-ionꢀNa]ꢀ
[C1-ionꢀNaꢀH2O]ꢀ and/or [B1-ionꢀNa]ꢀ
a
RI = relative intensity.
b
The different cleavages of the glycosidic bond are indicated according to the nomenclature of Domon and Costello.25
Table 3
Signals observed in the PSD spectra of compounds 1–4, 8, and 9
Compound
Matrix
DHB
Ion-mode
Positive
m/z (RI)a
Originb
1
347.7 (100%)
257.8 (6%)
[M+Na]+
0,3X1
2
DHB
Positive
508.0 (100%)
448.2 (5%)
346.9 (7%)
[M+Na]+
0,4X2 and/or 0,4X0
C2 and/or Y2
3
4
DHB
Positive
652.9 (11%)
289.8 (100%)
[M+Na]+
(B2 and/or Z2ꢀH2O)c
Nor-harmane
Negative
403.1 (100%)
259.1 (6%)
242.3 (17%)
403.1 (100%)
241.8 (56%)
[MꢀNa]ꢀ
C1
[C1ꢀH2O]ꢀ and/or B1
DHB
DHB
DHB
Negative
Negative
Negative
[MꢀNa]ꢀ
[C1ꢀH2O]ꢀ and/or B1
8
9
405.2 (100%)
242.3 (40%)
[MꢀNa]ꢀ
[C1ꢀH2O]ꢀ and/or B1
507.6 (100%)
405.2 (21%)
387.6 (21%)
507.8 (100%)
405.4 (16%)
387.1 (14%)
[MꢀNa]ꢀ
[MꢀNaSO3+HꢀNa]ꢀ
[MꢀNaSO3+HꢀNaꢀH2O]ꢀ
[MꢀNa]ꢀ
Nor-harmane
Negative
[MꢀNaSO3+HꢀNa]ꢀ
[MꢀNaSO3+HꢀNaꢀH2O]ꢀ
a
b
c
RI = relative intensity.
The different glycosidic and cross-ring cleavages are indicated according to the nomenclature of Domon and Costello.25
Protonated fragment.
together with a peak at m/z 387.9 due to dehydration of the latter
(Table 4). Compound 9 was also analyzed by using FA and the ion
liquid matrices FA–Bu and DHB–Bu. FA only yielded, the molecular
ion [MꢀNa]ꢀ with low intensity; on the other hand, FA–Bu gave in
linear and reflectron modes signals at m/z 507.7, 405.5 and 387.3
which were assigned as indicated in Table 4 for DHB. Finally by using
DHB–Bu, together with the signal at m/z 507.9, a large number of
high relative intensity signals of matrix clusters were observed.
Sulfated alditol 9 was also detected in positive-ion mode as
[M+Na]+ by using DHB and nor-harmane as matrices.
amounts of matrix clusters were formed. Nor-harmane yielded
signals at m/z 1018.6, 812.6, 710.0, 691.5, 504.7 and 402.9
assigned to [MꢀNa]ꢀ, [Mꢀ2NaSO3+2HꢀNa]ꢀ, [Mꢀ3NaSO3+
3HꢀNa]ꢀ, [Mꢀ3NaSO3+3HꢀNaꢀH2O]ꢀ, [C2-ionꢀNa]ꢀ and [C2-
ionꢀNaSO3+HꢀNa]ꢀ, respectively (Fig. 4). It is interesting to
point out that as it was previously observed in the MALDI MS
analysis of sulfated neocarrabiose oligosaccharides,19 the loss of
sulfate moieties from a molecular ion with x sulfate groups,
not necessarily produces the complete family of [MꢀnNaSO3+
nHꢀNa]ꢀ ions with n = 1, 2, 3, (x ꢀ 1). In the present example,
although the corresponding ions with n = 2 and n = 3 were ob-
served, the species [MꢀNaSO3+HꢀNa] was not detected. PSD
analysis was not successful because of the low intensity of the
[MꢀNa]ꢀ signal.
Finally, compound 10 (Table 1, Fig. 3) was analyzed by using
different matrices (Table 4). Although DHB and nor-harmane
yielded similar spectra in negative-ion mode, best results were
obtained when the latter was used as matrix, because minor

M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
279
Figure 2. Negative-ion mode MALDI-TOF mass spectrum of compound 5 obtained using nor-harmane as matrix. Peaks with an asterisk correspond to the matrix.
Figure 3. Molecular structure and cleavage of the glycosidic bonds in oligosaccharide alditols 6–10.
An additional and interesting conclusion is obtained by the
comparison of the relative intensities of the [MꢀNa]ꢀ and [MꢀnNa-
SO3+nHꢀNa]ꢀ signals for compounds 9 and 10. Compound 9, with
x = 2, shows an [MꢀNa]ꢀ to [MꢀNaSO3Na+HꢀNa]ꢀ intensity ratio
of 1.6:1, while for compound 10, with x = 4, the ratios of [MꢀNa]ꢀ
to [MꢀnNaSO3Na+HꢀNa] for n = 2 and 3 are 1.2:1 and 0.2:1,
respectively. Thus, the stability of the molecular ion is minor when
higher is the number of sulfate groups in the molecular structure of
a sulfated oligosaccharide alditol. Thus, both matrices DHB and
nor-harmane seem to yield molecular ions with enough vibrational

280
M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
Table 4
Molecular and daughter ions observed in the MALDI-TOF mass spectra of oligosaccharide alditols 6–10
Compound
Matrix
DHB
Ion-mode
Positive
Positive
Positive
m/z (RI)a
Assignmentb
[M+Na]+
[M+Na]+
6
7
8
349.5 (100%)
363.6 (100%)
DHB
DHB
451.4 (63%)
286.4 (100%)
[M+Na]+
[C1-ionꢀH2O+Na]+ and/or
[B1-ion+Na]+
Negative
405.2 (100%)
[MꢀNa]ꢀ
9
DHB
Positive
Negative
553.2 (100%)
507.6 (43%)
405.6 (60%)
387.9 (100%)
553.7 (100%)
507.8 (52%)
405.5 (33%)
248.7 (100%)
226.5 (98%)
[M+Na]+
[MꢀNa]ꢀ
[MꢀNaSO3+HꢀNa]ꢀ
[MꢀNaSO3+HꢀNaꢀH2O]ꢀ
[M+Na]+
Nor-harmane
Positive
Negative
[MꢀNa]ꢀ
[MꢀNaSO3+HꢀNa]ꢀ
[Z1-ionꢀH]ꢀ
[Z1-ionꢀNa]ꢀ
10
Nor-harmane
Negative
1018.6 (20%)
812.6 (17%)
710.0 (100%)
691.5 (15%)
504.7 (3%)
[MꢀNa]ꢀ
[Mꢀ2NaSO3+2HꢀNa]ꢀ
[Mꢀ3NaSO3+3HꢀNa]ꢀ
[Mꢀ3NaSO3+3HꢀNaꢀH2O]ꢀ
[C2-ionꢀNa]ꢀ
402.9 (37%)
[C2-ionꢀNaSO3+HꢀNa]ꢀ
a
RI = relative intensity.
b
The different cleavages of the glycosidic bond are indicated according to the nomenclature of Domon and Costello.25
Figure 4. Negative-ion mode MALDI-TOF mass spectrum of compound 10 obtained using nor-harmane as matrix. Peaks with an asterisk correspond to the matrix.
energy to suffer prompt fragmentation, mainly desulfation. Similar
results were obtained with the other MALDI matrices attempted
(FA, DHB–Bu and FA–Bu).
ative-ion mode, indicated that glycosidic C-cleavages and desulf-
ation were the main processes. The presence of the aldehyde as
terminal functional group in this sulfated oligosaccharide dimin-
ishes its thermal stability and the glycosidic C-cleavages efficiently
decompose the molecular ion.
3. Conclusions
It is important to note that this behavior is completely different
to that observed in the MALDI-TOF linear and reflectron spectra of
sulfated neocarrabiose oligosaccharides using DHB and nor-har-
mane as matrices where desulfation but no glycosidic C-cleavages
was detected as prompt fragmentation.19
Furthermore, monosulfated compounds 4 and 8 gave, in the
negative-ion PSD experiments (Table 3), cleavages of the glyco-
sidic bonds similarly to what was observed for the sulfated neo-
carrabiose oligosaccharides.19 On the other hand, for compound 9
desulfation was observed. PSD spectra of non-sulfated oligosac-
charides 1 and 2 showed, in the positive-ion mode cross-ring
fragmentation; for compound 2 a glycosidic cleavage was also
detected.
The results previously discussed showed that non-sulfated oli-
gosaccharides 1–3 and non-sulfated oligosaccharide alditols 6
and 7, which could only be analyzed in positive-ion mode
([M+Na]+), did not suffer any prompt fragmentation. On the other
hand, sulfated compounds 4 and 8–10 gave in the negative-ion
mode the molecular ion as [MꢀNa]ꢀ and compounds 4, 9 and 10
also showed prompt fragmentation (Tables 2 and 4). Compound
5, the G4S?DA?G4S?DA studied, had a quite peculiar behavior
because the intact molecular ion as [MꢀNa]ꢀ or [M+Na]+ was not
detected in the corresponding ion mode, although experiments
were conducted with different matrices in order to overcome this
problem. The species yielded by prompt fragmentation in the neg-

M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
281
The experiments conducted by using several compounds as
MALDI matrices allowed finding suitable conditions for effective
analysis of the alditols and oligosaccharides studied. The results
obtained can be summarized as follows:
60-Sulfated and methylated agarose, obtained from G. domingen-
sis,29 was submitted to partial reductive hydrolysis as previously
described.10 The hydrolysate was dissolved in water and applied
on a Sephadex G-25 column (1.5 ꢂ 100 cm) and eluted with water.
The main fraction was concentrated under vacuum, dissolved in
water, applied on a BioGel P-2 column (1.5 ꢂ 100 cm), and eluted
with water. The second subfraction obtained was concentrated
and freeze-dried to give 62-O-methyl agarobiitol 7.33
(a) Alditols with more than one sulfate group suffered desulf-
ation. Glycosidic cleavages were also observed, yielding the
more abundant ions in the case of compound 9 when nor-
harmane was used as matrix.
(b) Sulfated oligosaccharides with more than one sulfate group
gave, as it was expected,19 desulfation, but glycosidic frag-
mentation was more important.
(c) Non-sulfated oligosaccharides and non-sulfated alditols did
not show any prompt fragmentation.
(d) Non-sulfated oligosaccharides yielded cross-ring fragmenta-
tion in the PSD mode.
60-Sulfated and methylated agarose, obtained from G. domingen-
sis,29 was submitted to partial reductive hydrolysis as previously
described.10 The hydrolysate was dissolved in water, applied on a
DEAE–Sephadex A-25 (Clꢀ) column (2.5 ꢂ 12 cm), eluted with
water and a continuous NaCl gradient (0–0.15 M). The main frac-
tion eluted with NaCl was concentrated and desalted by aqueous
elution on a BioGel P-2 column (1.5 ꢂ 100 cm) to give agarobiitol
62-sulfate 8.14
Iota-carrageenan, obtained from E. denticulatum,28 was submit-
ted to partial reductive hydrolysis as previously described.10 The
hydrolysate was dissolved in water and applied to a DEAE–Sephar-
ose CL-6B (CO32ꢀ) column (3.0 ꢂ 15 cm). Successive elutions were
performed with water and ammonium carbonate solutions. Carra-
biitol 21,42-disulfate 9 and carratetraitol 21,42,23,44-tetrasulfate 10
were eluted with 0.10 M and 0.25 M ammonium carbonate solu-
tion respectively.33
TLC, NMR spectroscopy, and ESI MS were used to control the
purity of the compounds and their characterization. The 1H NMR
spectra of compounds 1, 3–10 and the 13C NMR spectra of all the
compounds are available as electronic supplementary data.
4. Experimental
4.1. Materials
Agarose was obtained from Vetec Quimica (Rio de Janeiro, RJ,
Brazil). Kappa-carrageenan, iota-carrageenan, 60-sulfated and meth-
ylated agarose were extracted from Kappaphycus alvarezii,26,27 Euch-
euma denticulatum28, and Gracilaria domingensis,14,29 respectively, as
it was previously reported.
4.2. Preparation of oligosaccharides and oligosaccharide
alditols
4.3. Electrospray-ionization mass spectrometry
Commercial agarose (150 mg) was first dissolved in hot (ꢁ90 °C)
water (13.5 mL) and then 1 M TFA solution (1.5 mL) was added in
one portion. The resulting mixture was heated at 65 °C for 2 h,
cooled to room temperature, diluted with water (15 mL), and then
concentrated under vacuum. The residue was co-evaporated with
toluene three times to give a sirup. This material was dissolved in
water (2 mL), applied on a BioGel P-2 column (1.5 ꢂ 70 cm) and
eluted with water. Oligosaccharide detection was performed by
the phenol–H2SO4 acidmethod30 and TLC. TLC was carried outon Sil-
ica Gel 60 (2:2:1 BuOH–AcOH–H2O) with detection by charring 0.5%
orcinol in concd H2SO4–EtOH 1:20. The fractions obtained were con-
centrated and freeze-dried to give agarobiose 1 and agarotetraose 3.
Oligosaccharide G?LA?G (2) was prepared as previously described
by Yang et al.31 with modifications32: a 1% mixture of commercial
agarose in concentrated HCl was kept at room temperature for
30 min. The reaction was then neutralized by the addition of solid
NaHCO3, the resulting mixture was freeze-dried and the remaining
solid was submitted to three successive extractions with DMSO.
The combined extracts were treated with three volumes of CHCl3
to precipitate the oligosaccharides, which were separated by centri-
fugation. The pellet was then applied on a BioGel P-2 column and
eluted with water; the eluate was monitored by the phenol–H2SO4
method which evidenced a rich sugar fraction that gave compound 2.
Kappa-carrageenan was submitted to partial acid hydrolysis as
described for commercial agarose. The hydrolysate was dissolved
in water, applied on a Sephadex G-25 column (1.5 ꢂ 100 cm), and
eluted with water. The fractions obtained were concentrated and
freeze-dried to give kappa-carrabiose 4 and kappa-carratetraose 5.26
Agaran sulfate isolated from Acanthophora spicifera was submit-
ted to partial reductive hydrolysis as previously described.10 The
hydrolyzate was dissolved in water, applied to a DEAE–Sephadex
A-25 (Clꢀ) column (5 ꢂ 22 cm) and eluted with water and NaCl
aqueous solutions. The aqueous eluate was concentrated and de-
salted by gel filtration chromatography on a BioGel P-2 column
(1.5 ꢂ 100 cm) using water as eluant. The fraction obtained was
concentrated and freeze-dried to give agarobiitol 6.13
The ESI MS equipment used was a Micromass Quatro LC–MS/MS
triple quadrupole mass spectrometer. Compounds 1–8 were di-
luted in 7:3 CH3CN–H2O at 1 mg mLꢀ1 and experiments were con-
ducted as it has been described elsewhere.14,26 Compounds 9 and
10 (0.125 mg mLꢀ1) were injected in 75:25 MeOH–20 mM n-
pentylammonium formate solution.33
Compound 1: positive-ion mode, m/z 671.4 (14%; [2M+Na]+),
381.3 (8%), 365.0 (56%; [M+H2O+Na]+), and 347.1 (100%; [M+Na]+).26
Compound 2: positive-ion mode, m/z 524.9 (85%; [M+K]+) and
508.9 (100%; [M+Na]+).32 Compound 3: positive-ion mode, m/z
671.6 (57%; [M+H2O+Na]+) and 653.4 (100%; [M+Na]+).26 Compound
4: negative-ion mode, m/z 421.2 (45%; [M+H2OꢀNa]+), 403.2 (100%;
[MꢀNa]+), 259.1 (40%; [C1-ionꢀNa]ꢀ), and 241.2 (8%; [B1-ionꢀ
Na]ꢀ).26 Compound 5: negative-ion mode, m/z 421,0 (18%; [C2-ion+
H2OꢀNa]ꢀ), 403.0 (90%; [C2-ionꢀNa]ꢀ), 394.0 (100%; [Mꢀ2Na]2ꢀ),
and 241.1 (9%; [B1-ionꢀNa]ꢀ).26
Compound 6: positive-ion mode, m/z 349.3 (100%; [M+Na])+.13
Compound 7: positive-ion mode, m/z 363.4 (100%; [M+Na])+.33
Compound 8: negative-ion mode, m/z 405.1 (100%; [MꢀNa])ꢀ.14
Compound 9: negative-ion mode, m/z 572.5 (11%; [Mꢀ2Na+
C5H11NH3]ꢀ), 485.3 (12%; [Mꢀ2Na+H]ꢀ), and 242.1 (100%;
[Mꢀ2Na]2ꢀ).33 Compound 10: negative-ion mode, m/z 1212.1
(11%; [Mꢀ4Na+3C5H11NH3]ꢀ), 1125.5 (29%; [Mꢀ4Na+H+2C5H11
-
NH3]ꢀ), 1037.5 (22%; [Mꢀ4Na+2H+C5H11NH3]ꢀ), 951.3 (12%;
[Mꢀ4Na+3H]ꢀ), 562.4 (86%; [Mꢀ4Na+2C5H11NH3]2ꢀ), 518.8
(100%; [Mꢀ4Na+H+C5H11NH3]2ꢀ), 475.4 (53%; [Mꢀ4Na+2H]2ꢀ),
345.7 (16%; [Mꢀ4Na+C5H11NH3]3ꢀ), and 316.4 (47%; [Mꢀ4Na+
H]3ꢀ).33
4.4. MALDI-TOF MS
4.4.1. Matrix chemicals
The b-carboline (9H-pyrido[3,4-b]indole) nor-harmane, 2,5-dihy-
droxybenzoic acid (DHB, gentisic acid), 3-methoxy-4-hydroxy-
cinnamic acid (ferulic acid, FA), n-butylamine, and 3,5-dimethoxy-

282
M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
4-hydroxycinnamic acid (sinapinic acid, SA; used for protein cali-
brants) were obtained from Aldrich Chemical Co.
4.4.7. Spectrum calibration
Spectra in the linear and reflectron modes were calibrated by
the use of external calibration reagents: (a) commercial proteins
(angiotensin I and neurotensin) with SA as matrix in positive-
and negative-ion modes, and (b) caffeine, b-estradiol-3-sulfate-
4.4.2. Calibrant chemicals
Caffeine (MW 149.19); b-estradiol-3-sulfate-17-glucuronide
dipotassium salt (MW 604.75),
a-cyclodextrin (cyclomaltohexaose,
17-glucuronide dipotassium salt, a-, b- and c-cyclodextrins with
MW 972.9), b-cyclodextrin (cyclomaltoheptaose, MW 1135.0),
c
-
nor-harmane as matrix in positive- and in negative-ion modes.
cyclodextrin (cyclomaltooctaose, MW 1297.1); angiotensin I (MW
1296.49), and neurotensin (N6383, MW 1672.96) were purchased
from Sigma–Aldrich.
The Kratos Kompact calibration program was used.
Acknowledgments
4.4.3. Solvents
Theauthorsare indebtedto theJapanSocietyfor thePromotionof
Science (JSPS) for Scientific Research (Grant-in-Aid (S)—20228004),
National Research Council of Argentina (CONICET; PIP 5443), the
Agencia Nacional de Promoción Científica y Tecnológica (ANPCYT;
PICT06-0615), the University of Buenos Aires (UBA X072 and
X137), Fundação Araucária/CNPq-Brasil (PRONEX-Carboidratos),
and MCT/CNPq/MS-SCTIE-DECIT/CT-Saúde-Brasil (No. 554671/
2006-9) for financial support. A.S.C., R.E.B., and M.C.M. are Research
Members of CONICET (Argentina). M.D.N. and M.E.D. are Research
Members of the National Research Council of Brazil (CNPq).
D.R.B.D. is grateful for a doctoral scholarship and A.G.G. for a post-
doctoral fellowship (PDJ), both from CNPq-Brasil. MALDI-TOF MS
was performed as part of the Academic Agreement between Rosa
Erra-Balsells (FCEyN-UBA, Argentina) and Hiroshi Nonami (CA-EU,
Japan) with the facilities of the High Resolution Liquid Chromatogra-
phy–integrated Mass Spectrometer System of the United Graduated
School of Agricultural Sciences (Ehime University, Japan).
MeOH and CH3CN (Sigma–Aldrich HPLC grade) were used as
purchased without further purification. Water of very low conduc-
tivity (Milli Q grade; 56–59 nS/cm with PURIC-S, ORUGANO Co.,
Ltd, Tokyo, Japan) was used.
4.4.4. Instrument
Measurements were performed using a Shimadzu Kratos, Kom-
pact MALDI 4 (Pulsed Extraction) laser desorption time-of-flight
mass spectrometer (Shimadzu, Kyoto, Japan), equipped with a
pulsed nitrogen laser (kem = 337 nm; pulse width = 3 ns), tunable
pulsed-delayed extraction (PDE) and PSD (MS/MS device) modes,
as described elsewhere.22–24
Experiments were performed using firstly the full range setting
for laser firing position in order to select the optimal position for
data collection, and secondly fixing the laser firing position in the
sample sweet spots. The sample was irradiated just above the
threshold laser power for obtaining molecular ions and with higher
laser power for studying cluster formation. Thus, the irradiation
used for producing a mass spectrum was analyte dependent. Usu-
ally 50 spectra were accumulated. The sample was measured in the
linear and reflectron modes, in both positive- and negative-ion
modes.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carres.2009.10.009.
4.4.5. Probe supports
References
Stainless steel polished surface twenty-sample-slides were pur-
chased from Shimadzu Co., Japan (P/N 670–19109-01). Polished
surface slides were used in order to get better images for morpho-
logical analysis in a stereoscopic microscope (NIKON Optiphot,
Tokyo, Japan; magnification 400ꢂ) and a high-resolution digi-
tal microscope (Keyence VH-6300, Osaka, Japan; magnification
800ꢂ).
1. Painter, T. J. Algal Polysaccharides. In The Polysaccharides; Aspinall, G. O., Ed.;
Academic Press: New York, 1983; Vol. 2, pp 195–285.
2. Zablackis, E.; Vreeland, V.; Doboszewski, B.; Laetsch, W. M. J. Phycol. 1991, 27,
241–248.
3. Knutsen, S. H.; Mysladodski, D. E.; Larsen, B.; Usov, A. I. Bot. Mar. 1994, 37, 163–
169.
4. Hurtado-Ponce, A. Q. Bot. Mar. 1995, 38, 215–219.
5. Ohno, M.; Nang, H. Q.; Hirase, S. J. Appl. Phycol. 1996, 8, 431–437.
6. Lechat, H.; Amat, M.; Mazoyer, J.; Gallant, D. J.; Buleon, A.; Lahaye, M. J. Appl.
Phycol. 1997, 9, 565–572.
4.4.6. Sample preparation
The ionic DHB–Bu and FA–Bu liquid matrices were prepared as a
1:1 molar mixture of the carboxylic acid and n-butylamine in MeOH
which, after evaporation under vacuum, yielded a viscous honey-
like ionic liquid matrix. The DHB–Bu ionic liquid matrix has been
previously described for the analysis of carbohydrates.34,35
Matrix stock solutions were made by dissolving 2 mg of the se-
lected compound in 0.2 mL of 1:1 MeOH–H2O or in 0.2 mL of 2:3
CH3CN–H2O. Analyte solutions were freshly prepared by dissolving
the carbohydrates (1 mg) in pure water (0.5 mL). Best results were
afforded when the matrix solutions were prepared in MeOH–H2O.
To prepare the analyte-matrix deposit two methods were
used: Method A (thin-film layer method) and Method B (mixture
method).19 Best results were obtained using nor-harmane as ma-
trix followed by the adoption of Method B. In this method, the
analyte stock solution was mixed with the matrix solution in
7. Estevez, J. M.; Ciancia, M.; Cerezo, A. S. Carbohydr. Res. 2000, 325, 287–299.
8. Aguilan, J. T.; Broom, J. E.; Hemmingson, J. A.; Dayrit, F. M.; Montaño, M. N. E.;
Dancel, M. C. A.; Niñonuevo, M. R.; Furneaux, R. H. Bot. Mar. 2003, 46, 179–192.
9. Estevez, J. M.; Ciancia, M.; Cerezo, A. S. Carbohydr. Res. 2004, 339, 2575–2592.
10. Usov, A. I.; Elashvili, M. Bot. Mar. 1991, 34, 553–560.
11. Falshaw, R.; Furneaux, R. H. Carbohydr. Res. 1995, 269, 183–189.
12. Usov, A. I.; Elashvili, M. Russ. J. Bioorg. Chem. 1997, 23, 502–511.
13. Gonçalves, A. G.; Ducatti, D. R. B.; Duarte, M. E. R.; Noseda, M. D. Carbohydr. Res.
2002, 337, 2443–2453.
14. Gonçalves, A. G.; Ducatti, D. R. B.; Paranha, R. G.; Duarte, M. E. R.; Noseda, M. D.
Carbohydr. Res. 2005, 340, 2123–2134.
15. Kazlowski, B.; Pan, C. L.; Ko, Y. T. Carbohydr. Res. 2008, 343, 2443–2450.
16. Harvey, D. J. Mass Spectrom. Rev. 1999, 18, 349–451.
17. Harvey, D. J. Mass Spectrom. Rev. 2006, 25, 595–662.
18. Harvey, D. J. Mass Spectrom. Rev. 2008, 27, 125–201.
19. Fukuyama, Y.; Ciancia, M.; Nonami, H.; Cerezo, A. S.; Erra-Balsells, R.;
Matulewicz, M. C. Carbohydr. Res. 2002, 337, 1553–1562.
20. Ciancia, M.; Sato, Y.; Nonami, H.; Cerezo, A. S.; Erra-Balsells, R.; Matulewicz, M.
C. ARKIVOC 2005, xii, 319–331.
21. Fukuyama, Y.; Kolender, A. A.; Nishioka, M.; Nonami, H.; Matulewicz, M. C.; Erra-
Balsells, R.; Cerezo, A. S. Rapid. Commun. Mass Spectrom. 2005, 19, 349–358.
22. Nonami, H.; Fukui, S.; Erra-Balsells, R. J. Mass Spectrom. 1997, 32, 287–296.
23. Erra-Balsells, R.; Nonami, H. Environ. Control Biol. 2002, 40, 55–73.
24. Erra-Balsells, R.; Nonami, H. Environ. Control Biol. 2008, 46, 65–90.
25. Domon, B.; Costello, C. E. Glycoconjugate J. 1988, 5, 397–409.
26. Ducatti, D. R. B. Hidrólise Ácida de Galactanas Para Produção de
Oligossacarídeos e C-Glicosídeos Building Blocks: Semi-síntese de N-Alquil e
1:4 v/v ratio. A 0.5 lL aliquot of this analyte-matrix solution
was deposited onto the stainless steel probe tip and dried with
a stream of forced room temperature air. Then, an additional por-
tion of 0.5 lL was applied to the dried solid layer on the probe,
causing it to redissolve partially, and the solvent was removed
by blowing air.

M. Kaniz Fatema et al. / Carbohydrate Research 345 (2010) 275–283
283
C-Dihidropiridina Glicosídeos. Ph.D Thesis, Universidade Federal do Paraná:
Brazil, 2009, p 135.
27. Ducatti, D. R. B.; Massi, A.; Noseda, M. D.; Duarte, M. E. R.; Dondoni, A. Org.
Biomol. Chem. 2009, 7, 578–588.
31. Yang, B. Y.; Brand, J.; Montgomery, R. Carbohydr. Res. 2001, 331, 59–67.
32. Gonçalves, A. G. Modificações Químicas em Oligossacarídeos Derivados de
Polissacarídeos de Algas Marinhas. Ph.D Thesis, Universidade Federal do
Paraná: Brazil, 2005, p 162.
28. Viana, A. G.; Noseda, M. D.; Duarte, M. E. R.; Cerezo, A. S. Carbohydr. Polym.
2004, 58, 455–460.
29. Guimaraes, M.; Ascencio, S. D.; Viana, A. G.; Plastino, E. M.; Duarte, M. E. R.;
Noseda, M. D. Bot. Mar. 2007, 50, 314–317.
30. Dubois, M.; Gilles, K. A.; Hamilton, J. K.; Rebers, P. A.; Smith, F. Anal. Chem.
1956, 28, 350–356.
33. Ducatti, D. R. B. Oligossacarídeos Obtidos por Hidrólise Redutiva Parcial de
Polissacarídeos de Algas Vermelhas: Estudos de Ressonância Magnética
Nuclear. MSc. Thesis, Universidade Federal do Paraná: Brazil, 2005, p 113.
34. Mank, M.; Stahl, B.; Boehm, G. Anal. Chem. 2004, 76, 2938–2950.
35. Laremore, T. N.; Murugesan, S.; Park, T.-J.; Avci, F. Y.; Zagorevski, D. V.;
Linhardt, R. J. Anal. Chem. 2006, 78, 1774–1779.