Reactions of aminoguanidine with a-dicarbonyl compounds studied by electrospray ionization mass spectrometry

Article abstract of DOI:10.1255/ejms.1191

Aminoguanidine possesses extensive pharmacological properties. This drug is recognized as a powerful a-dicarbonyl scavenger. In order to better elucidate the reactivity of aminoguanidine with a-dicarbonyls, aminoguanidine was reacted with several aldehydic and diketonic a-dicarbonyls. Electrospray ionization mass spectrometry is a suitable technique to study chemical and biochemical processes and was selected for the purpose. In aminoguanidine reactions, triazines were detected and other compounds that have never been reported before were identified. Triazine precursor forms were detected, namely tetrahydrotriazines and singly dehydrated tetrahydrotriazines. Moreover, species with bicyclic ring structures and dehydrated forms were also identified in aminoguanidine reactions. These species appear to result from tetrahydrotriazines and triazines reactions with one dicarbonyl molecule. Experiments revealed that these bicyclic species, in particular the ones resulting from triazines reactivity, could exist in solution, since they were both identified in the reactions of aminoguanidine and of a selected triazine with the dicarbonyls studied. The results obtained with regard to aminoguanidine/triazines reactivities appear to support the capability of triazines to condensate and form polycyclic ring structures and also to support literature mechanistic data for dihydroimidazotriazines formation via dihydroxyimidazolidine-triazines. The data obtained in this study may prove to be valuable to complement solution information concerning the reactivity of amines with a-dicarbonyls, in particular.

Full text of DOI:10.1255/ejms.1191

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
Received: 3 April 2012 Revised: 25 July 2012 Accepted: 26 July 2012 Publication: 6 September 2012
385
n
n
n
EUROPEAN
JOURNAL
OF
MASS
SPECTROMETRY
Reactions of aminoguanidine with
a-dicarbonyl compounds studied by
electrospray ionization mass spectrometry
Marco A. Saraiva,a–c,* Carlos M. Borgesa,b and M. Helena Florêncioa,b
aDepartment of Chemistry and Biochemistry, Faculty of Sciences of the University of Lisbon, Campo Grande, C8, 1749-016 Lisbon, Portugal.
E-mail: masaraiva@fc.ul.pt
bCenter for Chemistry and Biochemistry, Faculty of Sciences of the University of Lisbon, Campo Grande, C8, 1749-016 Lisbon, Portugal
cPresent address: Centro de Química Estrutural, Instituto Superior Técnico, Campus Alameda, Av. Rovisco Pais, Torre Sul, 1049-001 Lisbon,
Portugal
Aminoguanidine possesses extensive pharmacological properties. This drug is recognized as a powerful a-dicarbonyl scavenger. In
order to better elucidate the reactivity of aminoguanidine with a-dicarbonyls, aminoguanidine was reacted with several aldehydic
and diketonic a-dicarbonyls. Electrospray ionization mass spectrometry is a suitable technique to study chemical and biochemical
processes and was selected for the purpose. In aminoguanidine reactions, triazines were detected and other compounds that have
never been reported before were identified. Triazine precursor forms were detected, namely tetrahydrotriazines and singly dehydrated
tetrahydrotriazines. Moreover, species with bicyclic ring structures and dehydrated forms were also identified in aminoguanidine reac-
tions. These species appear to result from tetrahydrotriazines and triazines reactions with one dicarbonyl molecule. Experiments
revealed that these bicyclic species, in particular the ones resulting from triazines reactivity, could exist in solution, since they were
both identified in the reactions of aminoguanidine and of a selected triazine with the dicarbonyls studied. The results obtained with
regard to aminoguanidine/triazines reactivities appear to support the capability of triazines to condensate and form polycyclic ring
structures and also to support literature mechanistic data for dihydroimidazotriazines formation via dihydroxyimidazolidine-triazines.
The data obtained in this study may prove to be valuable to complement solution information concerning the reactivity of amines with
a-dicarbonyls, in particular.
Keywords: aminoguanidine, a-dicarbonyls, electrospray ionization mass spectrometry, aminotriazines
Introduction
The importance of the use of aminoguanidine in physiology and
pharmacology has been widely recognized.1–5 Aminoguanidine,
a nucleophilic hydrazine compound, contains a terminal amino
group with chemical reactivity higher than that of protein
terminal amino groups.5 Aminoguanidine has thus been
considered to be an effective compound for trapping reac-
tive carbonyl intermediates. In addition, aminoguanidine has
been extensively used in the control and inhibition of glycation
processes.1–5 Furthermore, the basic guanidinium functional
group is commonly used by proteins and enzymes as an essen-
tial recognition unit or catalytic moiety.6 Aminoguanidine7 is a
drug known to possess extensive pharmacological properties,
also serving as an important starting material for the synthesis
of drugs and enzyme inhibitors.6 Moreover, aminoguanidine has
been shown to possess pro-oxidant activity in vitro.8–10 It is also
able to generate hydrogen peroxide8 and superoxide anions,11 to
have chelating activity12 and to inactivate nitric oxide synthase
and semicarbazide-sensitive amine oxidase (SSAO).6
In this study, we decided to investigate the reactions of
aminoguanidine with selected a-dicarbonyls (Figure 1) by using
ISSN: 1469-0667
doi: 10.1255/ejms.1191
© IM Publications LLP 2012
All rights reserved

Reactions of Aminoguanidine with Dicarbonyls
386
electrospray ionization mass spectrometry (ESI-MS) tech-
niques. ESI-MS rapidly became a major analytical technique in
many branches of science and the number of ESI-MS applica-
tions is still growing steadily.13,14 A novel and successful appli-
cation of ESI-MS concerns the study of reaction mechanisms in
solution.13,14 Eberlin and co-workers reported a detailed over-
view of several important chemical reactions, accomplished
via gas-phase ion interception, characterization and reactivity
investigation of its key players.13,14 In a previous study,15 the
solutions of eight analytes (analytes with different acid/base
chemistry) were investigated under ESI-MS conditions, in the
presence of the 4-(2-hydroxyethyl)-1-piperazineethanesulfonic
acid (HEPES) buffer. The effect of solution parameters such as
analyte and HEPES buffer (electrolyte) concentrations, solu-
tion pH and analyte acid/base chemistry (pKa) on ESI-MS ion
response was investigated in that study.15 One of the analytes
investigated was the triazine, 3-amino-1,2,4-triazine, identi-
fied in the reaction of aminoguanidine with glyoxal.16,17 For
triazine solutions with a 10–3 M HEPES buffer, and similarly for
the other analytes selected, the analyte ESI mass spectra ion
responses varied linearly with their corresponding concen-
trations in solution, particularly for analyte concentrations
in solution above 10–5 M (i.e. generally in the average analyte
concentration range: 5×10–3 M–10–5 /10–6 M).15 Although high
Aldehydic -dicarbonyls
O
O
O
O
Glyoxal
Methylglyoxal
Diketonic -dicarbonyls
O
O
O
O
O
O
2,3-Hexanedione
Diacetyl
2,3-Pentanedione
O
O
O
O
3,4-Hexanedione
1-Phenyl-1,2-propanedione
Amines
NH
N
H2N
NH2
H2N
N
H
N
N
Aminoguanidine
3-amino-1,2,4-triazine
Figure 1. a-Dicarbonyls and amine compounds chemical structures.

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
387
amounts of a non-volatile buffer such as HEPES were used
in the analyte solutions to be electrosprayed, no significant
analyte ion signal suppression was observed. In fact, electro-
spraying these analyte solutions appeared to improve the ioni-
zation of analytes, especially for high analyte concentrations in
solution.15 These results are consistent with the ones obtained
by Constantopoulos et al.18 regarding the presence of impor-
tant amounts of electrolyte species in the analyte solutions and
analyte ion ESI response enhancement throughout the appli-
cation of the partitioning equilibrium model. Hence, the potent
capabilities of ESI-MS combined with the HEPES buffer ions
effect on the ionization of basic analytes can provide a suitable
methodology to investigate the reactivity of aminoguanidine
with the selected a-dicarbonyls. Furthermore, at solution
physiological pH aminoguanidine exists in the protonated form
and reactant and reaction product molecules (formed in the
reaction of aminoguanidine with a-dicarbonyls) can easily be
protonated and successfully transferred as isolated species
to the gas phase. This study aims to better elucidate amino-
guanidine reaction mechanistics and also to explore how the
were prepared by dissolving the compounds individually in a
previously prepared HEPES buffer solution (400mM). These
solutions pH was then adjusted to 7.5. Subsequently, 1.5mL
of the buffered aminoguanidine solution was added to 1.5mL
of the dicarbonyl buffered solution, in a total volume of 3mL.
After adding the dicarbonyl compounds, the solutions were
maintained incubated at 70°C, in a stove (Memmert, model
1500), under air and in the dark for a period of 28 days.
Dicarbonyls 1-phenyl-1,2-propanedione, 2,3-hexanedione
and 3,4-hexanedione solutions were of methanol/water
(60/40: v/v), accounting for their lack of solubility in water.
Moreover, in order to avoid reversible acetal formation for
the methylglyoxal dicarbonyl starting solution, the reaction
mixture was initiated at room temperature and subsequently
incubated at the desired temperature (70ºC). Time counting
for this latter reaction was not started before the first 30min
of reaction. Furthermore, aliquots of 200 mL were taken at
fixed time intervals, and immediately frozen at –80ºC (short
time periods).
dicarbonyl functionality affects aminoguanidine reactions and ESI-MS and ESI-MSn analysis
reaction products formation.
The mass spectra were obtained on a Finnigan LCQ Duo
ion trap mass spectrometer equipped with an electrospray
ionization (ESI) source. The source was maintained at a
4.5 kV voltage and coaxial sheath and auxiliary gases (both
nitrogen) flow-rates were ca 20–40 psi. Furthermore, the
capillary was kept at ca 10V voltage and maintained heated
at 220°C. The pressure (capillary/skimmer region), meas-
ured with the convectron gauge during electrospray experi-
ments, was normally 0.92torr. The base pressure, in the mass
analyser region, was ca 1.12 × 10–5 torr. The mass spectra
recorded were obtained by diluting stored samples 200 times
in ultrapure water. Samples were introduced directly in the
ESI source at a flow-rate of 5mLmin–1. Positive ion mode was
selected for the experiments carried out, and the mass range
was m/z 50–500. The mass spectrometer was operated using
three microscans with a maximum ion injection time of 50ms
(default values) and the recorded mass spectra were based in
one minute acquisition.
Experimental
Materials and methods
Materials
The amino compound aminoguanidine hydrochloride, the
dicarbonyls glyoxal, diacetyl, 1-phenyl-1,2-propanedione,
2,3-pentanedione, 2,3-hexanedione, 3,4-hexanedione, the
sulfonic buffer 4-(-2-hydroxyethyl)-1-piperazineethanesulfonic
acid sodium salt (HEPES sodium salt), the methylglyoxal dicar-
bonyl precursor methylglyoxal 1,1¢-dimethylacetal and the
triazine 3-amino-1,2,4-triazine, were purchased from Sigma
Chemical (JMGS, Lisbon). The organic solvent methanol (HPLC
p.a.) was from Merck (JMGS, Lisbon). Ultrapure deionised water
(18.2MWcm), generated in a Milli-Q plus system, was also used
in the preparation of the reaction mixtures. Concentrated solu-
tions of sodium hydroxide (NaOH) (solid pellets from Merck,
JMGS, Lisbon) and hydrochloric acid (HCl) (Riedel-de Haën,
JMGS, Lisbon) were used in order to adjust media pH. All chem-
icals used were of the highest quality available.
In the MS2 experiments, the selected precursor ions were
isolated in the ion trap and forced to collide with helium
gas. For ionic species excitation, collision energy, 20–50% of
the maximum available collision energy, was applied. This
amount of energy was chosen in order to ensure that the
precursor ions could be observed in the ESI-MS2 spectra, with
a remaining peak intensity of about 1–5% for precursor ions.
The mass spectrometer was operated using three microscans
with a maximum ion injection time of 200ms (default values)
and the recorded mass spectra were based in one minute
acquisitions. Total ion current was above 5×103 in all cases.
Furthermore, some precursor ions with low m/z values had
to be fragmented by setting the qz-excit parameter at different
magnitude orders (from qz-excit =0.25 to qz-excit =0.4). MSn exper-
iments were also carried out, especially for the higher molec-
ular weight precursor ions.
Chemical synthesis of methylglyoxal
In order to avoid some of the main contaminants existing in the
commercially methylglyoxal stock solutions available, methyl-
glyoxal was synthesised via methylglyoxal 1,1¢-dimethylacetal
acid hydrolysis (methylglyoxal yield ca 50%, as described in a
previous report).19
Preparation and incubation of the reaction
mixtures
In the aminoguanidine reaction mixtures preparation, both
the aminoguanidine and the dicarbonyls glyoxal, methylgly-
oxal, diacetyl, 1-phenyl-1,2-propanedione, 2,3-pentanedione,
2,3-hexanedione and 3,4-hexanedione solutions (200 mM)
Mass spectrometric data were obtained by using Xcallibur
version supporting software.

Reactions of Aminoguanidine with Dicarbonyls
388
x 3
261
x 3
x 3
261
261
100
100
80
60
40
20
0
100
80
60
40
80
60
40
20
0
97
20
97
0
133
97
50
100
150
200
m/z
250
300
350
400
50
100
150
200
250
300
350
400
50
100
150
200
250
300
350
400
(A1)
(A2)
(A3)
m/z
m/z
x 3
261
x 3
261
100
80
60
40
20
0
100
80
60
40
20
0
185
165
165
201
111
129
75
111
201
200
50
100
150
200
m/z
250
300
350
400
50
100
150
250
300
350
400
(B2)
(B3)
m/z
153
153
100
80
60
40
20
0
100
80
100
80
60
40
20
0
261
267
285
261
267
60
261
171
171
249
285
303
189
153
150
40
20
75
249
250
75
0
50
100
200
m/z
250
300
350
400
50
100
150
200
250
300
350
400
50
100
150
200
300
350
400
(C1)
(C2)
(C3)
m/z
m/z
Figure 2. ESI mass spectra of aliquots of glyoxal (A), methylglyoxal (B) and 2,3-hexanedione (C) reaction systems solutions. A1, B1 and
C1: 5 min; A2, B2 and C2: 1 day; A3, B3 and C3: 7 days. In the ESI mass spectra, the peaks for which m/z values are indicated, corre-
spond to [M+H]+ ions. The ions at m/z 75 (B1, B2, C1 and C2), 185 (B1 and B2) and 261 (all representations) correspond to the proto-
nated molecules of aminoguanidine, triazine+aminoguanidine adduct and HEPES buffer (sodium salt), respectively.
Reaction mixtures and ESI-MS analysis were performed
in duplicate. ESI-MSn spectra were repeated two to three
times.
dicarbonyl reactions were the ones that revealed more diver-
sity of ions in the ESI mass spectra.
Identification and characterization of
reaction products
ESI-MSn analysis data enabled us to establish ion structures
for the ions of interest (Tables 1 and 2), as shown in Tables 3 and
4. These attributes are sustained by the fragment ion patterns
depicted in Schemes 1–5. In order to make the discussion
on the ion structure attributes clearer, the ions of interest
are grouped in three main classes (a–c). These ion classes
encompass: ions 1a, 2a and 3a (class a); ions 1b and 2b (class
b); ions 1c and 2c (class c). The ion classes were based on
structural similarities, i.e. the ion classes encompass the
more structurally related ions. In addition, the more hydrated
the ions structures are, the more coherent are the fragmenta-
tion mechanisms for the dicarbonyl reaction systems studied.
When the ions structures are less hydrated, the more diversity
in fragmentation mechanisms occurs, which is dependent on
the stability of the ions structures and the nucleophile/electro-
phile properties of the dicarbonyl substituents.
Results and discussion
Behavior of dicarbonyl reaction systems
under ESI mass spectrometry conditions
One of the advantages of ESI-MS is to enable the detection of
pre-existing ions in solution. The reactions of aminoguanidine
with dicarbonyls glyoxal, methylglyoxal, diacetyl, 1-phenyl-
1,2-propanedione, 2,3-pentanedione, 2,3-hexanedione and
3,4-hexanedione were studied in this work. ESI mass spectra of
aliquots of reaction system solutions were obtained, revealing
little or no fragmentation, as it can be seen, for example,
for glyoxal, methylglyoxal and 2,3-hexanedione dicarbonyl
reaction systems (Figure 2). The ESI mass spectra exhibited
little or no fragmentation, meaning that the ESI-MS condi-
tions used were suitable for the species ionization and ion
generation in the gas phase. The most intense peaks, apart
from the ones corresponding to the HEPES buffer system,
were attributed to protonated reaction product molecules. Class a (ions 1a–3a)
Relative ion abundances are presented in Tables 1 and 2 for
all reaction systems. Furthermore, the ion abundances were
normalized to m/z 261 buffer ion abundances and determined
at three different reaction times. Moreover, with respect to
the dicarbonyl reactions studied, methylglyoxal and diketonic
Fragmentation of ions 1a, at m/z 133 (glyoxal), 161 (diacetyl),
223 (1-phenyl-1,2-propanedione) (Table 1), 175 (2,3-pentan-
edione) and 189 (2,3-hexanedione and 3,4-hexanedione) (Table
2), are depicted in Scheme 1. These ions fragmented to give
prominent losses of one water molecule and one dicarbonyl

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
389

Reactions of Aminoguanidine with Dicarbonyls
390
H+
H+
H+
H+
R1
R1
R1
R1
H
N
H
N
H
N
N
N
N
- H2O
N
- H3N
HO
N
O
O
O
HO
N
N
H
NH2
N
H
NH2
N
H
NH2
R2
R2
R2
R2
- CH2N4
- H2O
- dicarb
H+
- CH2N2
R1
O
H+
H+
R1
N
N
H2N
H+
N
N
R2
R1
(2,3-PD, 2,3-HD and 3,4-HD)
NH
NH
R2
NH2
H2N
NH2
O
Scheme 1. Fragmentation mechanism of ions 1a. R1 =H
R2
(glyoxal and methylglyoxal), CH3 (diacetyl, 2,3-pentanedione,
2,3-hexanedione and 1-phenyl-1,2-propanedione) and CH2CH3
(3,4-hexanedione). R2 =H (glyoxal), CH3 (methylglyoxal and
diacetyl), CH2CH3 (2,3-pentanedione and 3,4-hexanedione),
CH2CH2CH3 (2,3-hexanedione) and phenyl (1-phenyl-1,2-pro-
panedione).
Scheme 2. Fragmentation mechanism of ions 2a. R1 and R2 as
in Scheme 1.
molecules. The dihemiaminal (double hemiaminal), in dihy-
droxyimidazolidines structures, is not very stable with respect
to water elimination, which normally spontaneously occurs in
amine-carbonyl reactions. Indeed, the stability of hemiami-
nals decreases with the presence of electron-donor substitu-
ents, although it increases with increasing electronegativity of
the substituents. Therefore, more stable hemiaminals occurs
for glyoxal- and 1-phenyl-1,2-propanedione-derived tetra-
hydrotriazines. Upon fragmentation, ions 1a for glyoxal and
1-phenylpropanedione easily lose one water molecule (more
stable tetrahydrotriazines) but the same is not expected for
diacetyl. Thus, ions 1a for diacetyl will probably lose one dicar-
bonyl molecule, as observed in the main fragmentation mech-
anism (Scheme 1).
Ions 2a, at m/z 129 (methylglyoxal), 205 (1-phenyl-
1,2-propanedione) (Table 1), 157 (2,3-pentanedione) and
171 (2,3-hexanedione and 3,4-hexanedione) (Table 2), frag-
mented to give a more complex fragment ion composition,
in comparison to ions 1a (scheme 2). Losses of one water
molecule and CH2N2 were the most prominent ones observed
in ions 2a fragmentation. Loss of CH2N4 was observed for
the diketonic dicarbonyl reaction systems only (Scheme 2),
appearing to result from a hetero ring structure and not from
an open chain aminoguanidine-derived molecule. In fact,
for ions 2a to possess an open chain structure, the occur-
rence of the mentioned CH2N4 loss required a long range
proton displacement, which is not favorable in comparison to
what is predicted for the fragmentation of ions 2a with cyclic
structures. The loss of CH2N2, in ions 2a fragmentation, also
observed in ions 1a fragmentation, after the elimination of
one water molecule, suggests that ion 1a and 2a structures
are related, although the latter ion structures should be more
stable than the former. These observations led us to attribute
the structures of protonated tetrahydrotriazine–H2O mole-
cules to ions 2a (Table 3). It is understood that, after ions 2a
lose CH2N2, substituted diazetidine ions appears as fragment
molecule (Scheme 1). After ions 1a lose one dicarbonyl mole-
cule, protonated aminoguanidine molecules result as frag-
ment ions. The structure of protonated tetrahydrotriazine
molecules was attributed to ions 1a (Table 3). Ions 1a fragment
ion composition was found to be similar to the ones proposed
for acetyl-dihydroxyimidazolidines20 (resulting from acetyl-
arginine reactions) and dihydroxyimidazolidines21 (resulting
from guanidine reactions), where water and dicarbonyl mole-
cules losses mainly occurred. This observation reinforces
the ion structures attributed. With respect to ions 1a frag-
mentation, protonated molecules generated from glyoxal and
1-phenyl-1,2-propanedione show a prominent loss of water
while for diacetyl the prominent loss was dicarbonyl. This
difference in the fragmentation mechanisms is explained by
the electrophile/nucleophile nature of the reacting dicarbonyl
H+
H+
R1
R1
N
N
N
N
- CHN
N
NH
R2
R2
NH2
- CH2N2
H+
R1
R2
R1'
R2'
R1
R2
N
N
N
N
N
N
(2,3-PD, 2,3-HD and 3,4-HD)
Scheme 3. Framentation mechanism of ions 3a. R1 and R2 as in
Scheme 1. R1¢=R1 –H and R2¢=R2 –H.

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
391
H+
H+
R2
R2
R1
R1
H
N
H
N
OH
OH
R1
OH
OH
R1
HO
HO
N
N
- H2O
O
N
H
N
N
H
N
R2
R2
(ions 2b structure by ESI-MS3 analysis)
ions 1b
H+
H+
H+
R2
R2
R2
H
H
H
OH
R1
O
R1'
N
R1
O
N
N
O
O
N
N
N
- H2O
- H2O
OH
R1
R1
R1
N
H
N
N
N
N
H
N
R2
R2'
R2
H
ions 2b
- dicarb
- CH2N2
H+
H+
R1
O
R1
H
N
R1
N
N
NH
O
O
R2
R2
N
H
NH
R2
(ions 2a structure by ESI-MS3 analysis)
Scheme 4. Fragmentation mechanisms of ions 1b and 2b. R1 and R2 as in Scheme 1. R1¢=R1 –H and R2¢=R2 –H.
H+
H+
R2
H+
R2
H
N
H
N
H
N
R1'
R2'
R1'
R2'
R1
O
O
O
- H2O
- CO
N
N
CH3
N
R1
R1
N
H
N
N
H
N
N
H
N
R2
(MG)
-CH2N2
ions 1c
H+
- CH2N2
R1'
R1
O
- C2H4
N
N
H+
R1
R1
R2'
R2
N
N
O
O
H+
H
N
R1
O
N
R2
R2
N
O
N
H
R2
(PPD)
H+
H+
R2
R2'
H
H
N
R1'
R2'
R1'
R2'
N
O
- H2O
N
N
R1'
R1
N
H
N
N
H
N
H+
ions 2c
- CO
H
N
R1'
- CH2N2
- C2H4
N
CH3
H+
N
H
N
R1'
R1
R2'
H+
(MG)
N
N
O
H
R1'
N
O
R2'
R2
N
N
H
N
R2'
(PPD)
Scheme 5. Fragmentation mechanisms of ions 1c and 2c. R1 and R2 as in Scheme 1. R1¢=R1 –H and R2¢=R2 –H.

Reactions of Aminoguanidine with Dicarbonyls
392
Table 3. Proposed ion structures for the compounds of interest identified in the ESI mass spectra; ions 1a–1b (Tables 1 and 2)
Reaction systems
Possible ion structuresa
tetrahydrotriazine
tetrahydrotriazine– H2O
triazine
(ions 3a)
dihydroxyimidazolidine-
tetrahydrotriazine
(ions 1b)
(ions 1a) b
(ions 2a) c
Glyoxal
—
—
Methylglyoxal
Diacetyl
1-Phenyl-1,2-
propanedione
2,3-Pentanedione
2,3-Hexanedione
3,4-Hexanedione
aAll ion structures proposed refer to singly protonated molecules, [M + H]+.
bIon structure not detected in the methylglyoxal reaction system.
cIon structure not detected in the diacetyl (and glyoxal) reaction system.
—Not detected.
R1 = H (glyoxal and methylglyoxal), CH3 (diacetyl, 2,3-pentanedione, 2,3-hexanedione and 1-phenyl-1,2-propanedione) and CH2CH3 (3,4-hexanedione).
R2 = H (glyoxal), CH3 (methylglyoxal and diacetyl), CH2CH3 (2,3-pentanedione and 3,4-hexanedione), CH2CH2CH3 (2,3-hexanedione) and
phenyl (1-phenyl-1,2-propanedione).
Table 4. Proposed ion structures for the compounds of interest identified in the ESI mass spectra; ions 2b–2c (Tables 1 and 2).
Reaction
systems
Possible ion structuresa
(dihydroxy-
imidazolidine-
tetrahydro-
triazine)–H2O
(ions 2b)
(dihydroxyimidazolidine-
tetrahydrotriazine)–2H2O(b) and
dihydroxyimidazolidine-triazine(c)
(ions 1c)
(dihydroxyimidazolidine-
tetrahydrotriazine)–3H2O
(ions 2c)
Glyoxal
–
–
–
Methylglyoxal
Diacetyl
1-Phenyl-1,2-
propanedione
(b)
2,3-Pentanedione
2,3-Hexanedione
R2
OH
R1
R2
N
N
N
OH
R1
N
3,4-Hexanedione
(c)
aAll ion structures proposed refer to singly protonated molecules, [M+H]+. Two alternative ion structures b,c were attributed to ions 1c. – Not detected.
R1 =H (methylglyoxal), CH3 (diacetyl, 2,3-pentanedione, 2,3-hexanedione and 1-phenyl-1,2-propanedione) and CH2CH3 (3,4-hexanedione).
R2 =CH3 (methylglyoxal and diacetyl), CH2CH3 (2,3-pentanedione and 3,4-hexanedione), CH2CH2CH3 (2,3-hexanedione) and
phenyl (1-phenyl-1,2-propanedione).

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
393
ions. This four-membered nitrogen heterocycle represents a
very strained ring system.
structures of protonated dihydroxyimidazolidine-tetrahydrotri-
azine molecules to ions 1b (Table 3).
Ions 3a, at m/z 97 (glyoxal), 111 (methylglyoxal), 125
(diacetyl), 187 (1-phenyl-1,2-propanedione) (Table 1), 139 Class c (ions 1c and 2c)
(2,3-pentanedione), 153 (2,3-hexanedione and 3,4-hexan-
edione) (Table 2), fragmented to give rise to somewhat
distinct fragment ion patterns among the reaction dicar-
bonyl systems studied. The losses of CHN and CH2N2 were
observed as the most common (Scheme 3). Since the promi-
nent CH2N2 loss was equally observed in ions 2a and 3a
fragment ion compositions, for the heavier alkyl-diketonic
dicarbonyls, it seems reasonable to assume that these ions
possess similar ion structures. Therefore, the structures
of protonated triazine molecules were attributed to ions 3a
(Table 3). It is to bear that a commercially available triazine,
with a molecular structure identical to the aminoguanidine
glyoxal-derived triazine, revealed, in their fragmentation, the
unique and important loss of CHN (not shown). This result
was identical to the observed in the fragmentation of the
protonated glyoxal-derived triazine molecules (under the
same fragmentation conditions; abundance of precursor
ions, collision energy and qz-excit adjusted to 0.4) (Table 3). It
deserves mentioning that the loss of CHN represents a loss
from very stable precursor ion structures, which reinforces
the ions structures attributed to ions 3a.
Ions 1c, at m/z 183 (methylglyoxal), 211 (diacetyl), 335
(1-phenyl-1,2-propanedione) (Table 1), 239 (2,3-pentane-
dione), 267 (2,3-hexanedione and 3,4-hexanedione) (Table 2),
revealed similar fragment ion compositions, where the most
abundant losses are concerned. Prominent losses are one
water molecule and CH2N2 (Scheme 5). Two alternative ion
structures were proposed for ions 1c, since the fragment ion
composition of ions 1c does not support unique ion struc-
ture attributes. Thus, significant losses of CH2N2, also occur-
ring after the loss of one water molecule, NH3, CO (methyl-
glyoxal) and C2H4 (1-phenyl-1,2-propandione), are indicative
of tetrahydrotriazines–H2O fragmentation (ions 2a) (Scheme
2). Therefore, the structures of protonated molecules of
(dihydroxyimidazolidine-tetrahydrotriazines)–2H2O have been
attributed to ions 1c (Table 4–c). A competitive fragmentation
mechanism also seems to occur. This means the occurrence of
one prominent water molecule loss, particularly for the alkyl-
diketonic dicarbonyl reaction systems (diacetyl, 2,3-pentan-
edione, 2,3-hexanedione and 3,4-hexanedione), suggesting an
imidazolidine ring development. Hence, the structures of the
protonated molecules of dihydroxyimidazolidine-triazine were
attributed to ions 1c also (Table 4–b). Moreover, ESI-MS3 anal-
ysis suggests that the fragment ions resulting from precursor
ions 2b, from one water molecule loss, should have the same
fragment ion composition as ions 1c for the alkyl-diketonic
dicarbonyls, diacetyl, 2,3-pentanedione, 2,3-hexanedione and
3,4-hexanedione, especially. Thus, we can assume that, for
the alkyl-diketonic dicarbonyls, the formation of (dihydroxy-
imidazolidine-tetrahydrotriazines)–2H2O could prevail over the
formation of dihydroxyimidazolidine-triazines.
Ions 2c, at m/z 165 (methylglyoxal), 193 (diacetyl), 317
(1-phenyl-1,2-propanedione) (Table 1), 221 (2,3-pentane-
dione), 249 (2,3-hexanedione) (Table 2), fragmented to give, in
general, one water molecule and CH2N2 as the major losses
(Scheme 5). The fact that, in ions 2c fragmentation, neutral
losses appear to result from a triazine ring formed and from
the development of dicarbonyl moieties in a hetero ring
structure (CO and C2H4 losses), suggesting the occurrence
of imidazole ring formation, connected to a triazine ring in
ions 2c structures. Thus, we attributed the structures of the
protonated (dihydroxyimidazolidine-tetrahydrotriazine)–3H2O
molecules to ions 2c (Table 4).
Class b (ions 1b and 2b)
Ions 1b, at m/z 191 (glyoxal), 219 (methylglyoxal), 247 (diacetyl),
371 (1-phenyl-1,2-propanedione) (Table 1), 275 (2,3-pentan-
edione) and 303 (2,3-hexanedione) (Table 2), fragmented
similarly, with losses of one to two water molecules, namely
Scheme 4. Loss of two water molecules also suggests that
two dicarbonyl molecules must be involved in ions 1b forma-
tion. This can be rationalized in terms of the water molecule
losses observed in ions 1a and 2a fragment ion compositions
where, in either ion structures, one water molecule loss was
observed. Furthermore, ESI-MS3 analysis of the fragment ions
derived from precursor ions 1b, by loss of one water mole-
cule, revealed a fragment ion composition similar to the one
observed for ions 2b, further discussed (Scheme 4). To better
explain the structural attributes of ions 1b, ions 2b should first
be discussed. Ions 2b, at m/z 201 (methylglyoxal), 229 (diacetyl),
353 (1-phenyl-1,2-propanedione) (Table 1), 257 (2,3-pentan-
edione), 285 (2,3-hexanedione and 3,4-hexanedione) (Table
2), fragment to give losses of one to two water molecules,
the second water molecule loss being more significant for
the heavier alkyl-diketonic dicarbonyls. An important loss of
one dicarbonyl molecule was also observed in the fragment Considerations regarding the reaction
ion composition of ions 2b (Scheme 4). Furthermore, ESI-MS3 products formed
analysis of the most abundant fragment ions proved that after
ions 2b lose one dicarbonyl molecule, ions 2a resulted as frag-
ment ions. Two dicarbonyl molecules are indeed involved in
ions 2b formation. This led us to attribute the structure of the
protonated (dihydroxyimidazolidine-tetrahydrotriazine)–H2O
molecules to ions 2b (Table 4). Moreover, knowing that ions 1b
fragment ions, resulting from the loss of one water molecule,
were indeed found to be ions 2b (ESI-MS3), we attributed the
In this section, it is important to refer to the fact that, according
to the methodology developed by us to study reactions in solu-
tion by ESI-MS off-line techniques,15 the relative abundances
of ions identified in the ESI mass spectra vary linearly with ion
concentrations in solution, especially for ions concentrations
in the following range 10–5–10–3 M.15 This concentration range
appears, therefore, to be suitable for studying the reactions in
question, since the HEPES buffer species appear to enhance

Reactions of Aminoguanidine with Dicarbonyls
394
ion ESI signals. Although comparisons among the relative
ion abundances obtained are allowed, such comparisons are
only considered for ions belonging to each class, i.e. the ions
more structurally related and with closer m/z values. This
minimizes different ionization efficiencies that analyte ions
can develop towards the HEPES buffer ions and minimizes
mass-dependent discrimination22 in the mass spectrometer.
In the reactions of aminoguanidine with the selected
a-dicarbonyls, the compounds belonging to two different
classes were identified, namely triazines (aminotriazines)
and their hydrated forms and compounds having bicyclic ring
structures. In the literature, triazines were first isolated by
Erickson,16 during the study of the reactions of aminoguani-
dine salts with a-dicarbonyl compounds such as glyoxal and
diacetyl. More recently, triazine compounds have been reported
to also be formed in the reaction of aminoguanidine with some
of the dicarbonyl compounds used here, in particular glyoxal
and methylglyoxal.17 In another study, the latter reactions
(including that of 3-deoxyglucosone) were investigated with
concern to their kinetics and mechanism, under physiological
conditions.23 Reactions kinetics was found to be different,
accounting for the reacted dicarbonyls and also dependent
on the hydrated and non-hydrated dicarbonyl forms present
in solution, as in the case of the methylglyoxal dicarbonyl.23 In
the present study, the aminoguanidine reactions were moni-
tored under ESI-MS conditions and triazines were detected in
the beginning of the reactions, ca five minutes (Tables 1 and
2). They were, however, even detected after 14 days reaction
(not shown). This observation is consistent with triazines being
relatively stable compounds, at least under the reaction condi-
tions investigated. Tetrahydrotriazines (1a) and tetrahydro-
triazines–H2O (2a) were also detected for most of the reaction
systems investigated. Their detection occurred mostly in the
beginning of the reactions, ca five minutes to one day (Tables 1
and 2), suggesting that they could be triazine reaction precur-
sors. In the literature, it has been observed that the reaction
of aminotriazine with a-halo carbonyl compounds, in partic-
ular, is regioselective, regarding triazines formation.24 To this
purpose, the reactions of aminoguanidine with compounds
having two electrophilic sites, with different energies, can
proceed to the formation of isomeric triazine reaction prod-
ucts.24 The formation of triazine isomeric reaction products
have also been observed in the reaction of aminoguanidine
with methylglyoxal.17 Limanto et al., in a study concerning a
regioselective approach development for substituted triazines
formation, have reasoned that the formation of hydrazone,
an intermediate involved in triazines formation, occurs when
the difference in reactivity between the two electrophilic sites
of a-halo carbonyls is more pronounced.24 Nevertheless, in
the present study, we did not identify any hydrazone form in
the reactions of aminoguanidine, which can be explained by
the minor different electrophile energetic difference of the
dicarbonyl molecules studied. In fact, gas-phase results indi-
cated that hydrated triazines forms exist preferably in cyclic
structures.
with diketonic dicarbonyls, it appears that tetrahydrotriazines
and triazines react with a subsequent dicarbonyl molecule.
This situation can indeed occur, especially for triazines, since
the heterocyclic structures in question can have several equi-
librium species, allowing, therefore, for the favorable forma-
tion of bicyclic compounds. The existence of structurally
different triazines in solution could be due to the p electron
density redistribution in aminoguanidine molecules, which
is strong, even when compared to its biogenic guanidine.12
In order to prove that the mentioned bicyclic ring structures
could be formed in the reaction of triazine (aminotriazine)
with a dicarbonyl molecules, we have reacted a commercially
available 3-amino-1,2,4-triazine (aminoguanidine glyoxal-
derived triazine) with diacetyl and 2,3-hexanedione, in partic-
ular. Several reaction products were identified under ESI-MS
conditions (not shown). These reaction products corresponded
mostly to the formation of bicyclic compounds, with molecular
structures similar to dihydroxyimidazolidine-triazines (1c)
and to their dehydrated forms, identified in aminoguanidine
reactions. The fragmentation of the corresponding proto-
nated molecules led us to assume that the formation of a
bicyclic ring can occur in solution and that the reacted dicar-
bonyl molecule should not be present in an open chain form.
Therefore, since dihydroxyimidazolidines–triazine, in partic-
ular, were identified in both the reactions of aminoguanidine
and of an aminotriazine with dicarbonyls, it is more likely that
these compounds are reaction products present in solution
and not species formed in the gas-phase environment of the
mass spectrometer. To reinforce this assumption, the forma-
tion of a dihydroxyimidazolidine–triazine derivative, resulting
from the reaction of a substituted aminotriazine with glyoxal
in the presence of alcoholic solvents, was reported in the
literature.25 This compound was also reported to be involved
in the reaction mechanism proposed for the formation of
dihydroimidazotriazines.24 In this study, Garnier et al. did not
succeed in isolating the dihydroxyimidazolidine–triazine inter-
mediate, but they were able to establish that the resulting
dehydrated form of the intermediate was indeed unstable.25
The mentioned dehydrated form corresponds to a single dehy-
drated dihydroxyimidazolidine-triazine derivative (Table 4). In
this study,25 it also deserves mentioning that the reaction of
substituted triazines with glyoxal, in the presence of alcohol,
was performed at 80°C, which resembles the reaction condi-
tions used in the present study (70°C). Similar to what has
been described by Garnier et al.,25 we were also unable to
identify a dihydroxyimidazolidine–triazine for the reaction of
aminoguanidine with glyoxal. Nevertheless, in the present
study, this compound was observed for the diketonic dicar-
bonyl systems, in particular. This observation may suggest
that more stable dihydroxyimidazolidine–triazines are prob-
ably developed in the diketonic dicarbonyl reaction systems, in
comparison to the aldehydic dicarbonyl systems, also further
reinforced by the significant formation of triazines in the dike-
tonic dicarbonyl systems. In the diketonic dicarbonyl systems,
both dihydroxyimidazolidine–triazine and their corresponding
dehydrated forms were identified. This observation seems to
agree with the reaction mechanism proposed by Garnier et
Regardingthebicyclicringstructuresformed,i.e.compounds
1b, 2b, 1c and 2c, mainly in the reactions of aminoguanidine

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
395
R1
HO
HO
R2
R1
H
N
H
N
R1
R2
N
- H2O
- H2O
N
N
N
O
N
H
NH2
N
H
NH2
N
NH2
R2
+ dicarbonyl
tetrahydrotriazine
tetrahydrotriazine - H2O
triazine
R2
+ dicarbonyl
OH
OH
R1
R1
N
N
R2
R2
R2
R1
R2
R1
R2
H
N
H
N
OH
OH
R1
OH
R2
N
N
R1
R2
N
N
O
HO
HO
- H2O
- H2O
dihydroxyimidazolidine-triazine
- H2O
N
N
N
OH
R1
O
+
R1
R2
R1
H
N
N
H
N
N
N
H
N
O
N
dihydroxyimidazolidine-tetrahydrotriazine - H2O
dihydroxyimidazolidine-tetrahydrotriazine
dihydroxyimidazolidine-tetrahydrotriazine - 3H2O
O
R1
N
H
N
R2
dihydroxyimidazolidine-tetrahydrotriazine - 2H2O
Scheme 6. Reaction mechanism proposed for the reaction aminoguanidine/selected triazine with the dicarbonyls. R1 and R2 as in
Scheme 1.
al. for the formation of dihydroimidazotriazines via dihydroxy-
imidazolidine–triazines.25 It also appears to be consistent with
the fact that dehydrated dihydroxyimidazolidine–triazines can
be generated in the reaction of triazines with a-dicarbonyls.
The results concerning the reactivity of aminoguanidine and
the triazine investigated, in particular, seem to further sustain
the ability and importance of triazines condensations, which
has been recognized as potentially active building blocks in
fields such as agrochemical and medicine.25,26 With respect
to the nature of the dicarbonyl electrophiles of the reacting
dicarbonyl molecules, the reaction of aminoguanidine with the
selected dicarbonyls seems to proceed to double condensa-
tion of dicarbonyls when the carbonyl electrophiles are more
stabilized, which is the exception for the glyoxal dicarbonyl.
Symmetry aspects of the dicarbonyl electrophiles do not seem
to have a marked influence on the aminoguanidine reactions.
In Scheme 6, the reaction mechanism proposed for the reac-
tion of aminoguanidine/selected triazine with the dicarbonyl
compounds is depicted.
with a-dicarbonyls seems, therefore, to be very complex. As
well as the reported formation of triazines already proposed
to occur in solution, for the reaction of aminoguanidine with
some dicarbonyls here studied, for example, glyoxal, methyl-
glyoxal and diacetyl,16,17 possible triazines precursors such as
tetrahydrotriazines and singly dehydrated tetrahydrotriazines,
were identified for most of the reaction systems investigated.
Bicyclic ring structures appear to result from the reaction of
tetrahydrotriazines and triazines with one subsequent dicar-
bonyl molecule. These bicyclic ring structures were identified
in the reactions of aminoguanidine and of a selected triazine
with the dicarbonyls studied, suggesting that they could be
reaction products formed in solution. Concerning the reactivity
of triazines, in particular, the possible formation of triazine
condensation products, in the reaction of substituted triazines
with glyoxal has been reported in the literature.25 Despite
the fact that these condensation products were not identified
in the literature due to their intrinsic instability, they were
included as part of the reaction mechanism proposed for the
formation of dihydroimidazotriazines. The latter compounds
are structurally related to the dihydroxyimidazolidine-triazines
identified in the present study. These latter compounds were
also not observed in the reactions of aminoguanidine/selected
triazine with glyoxal. Nonetheless, for diketonic dicarbonyl
systems, in particular, the formation of more stable dihydrox-
yimidazolidine-triazines eventually resulted, along with its
dehydrated forms, supporting the fact that these species can,
indeed, be formed in the reaction of aminoguanidine/selected
triazine with a-dicarbonyls. This also supports the reaction
mechanism proposed for the formation of dihydroimidazo-
triazines via dihydroxyimidazolidine-triazines. The present
results regarding aminoguanidine/triazine reactivities further
reinforce the importance and versatility of triazines conden-
sations in the generation of polycyclic ring structures, which
have been found to be of recognized use in fields such as
agrochemical and medicine.26,27
Conclusions
The reactions of aminoguanidine with several aldehydic and
diketonic a-dicarbonyls were studied and reaction products
identified. Within each identified ions class, the most dehy-
drated compound forms revealed losses that can support
the formation of stable heterocyclic ring structures. Losses
from the dicarbonyl moieties formed, including dicarbonyl
substituents and carbonyl/epoxy group formation, were
observed in the fragmentation of most dehydrated ion struc-
tures identified. In each compound class assigned, the ions
differed from 18Da and the fragment ion compositions were
consistent with increased stability of the most dehydrated
ion forms, in comparison to their precursor forms, within
the class of ions assigned. The reaction of aminoguanidine

Reactions of Aminoguanidine with Dicarbonyls
396
This study provides information on the aminoguanidine/
selected triazine reactivity with a-dicarbonyls. This infor-
mation can be valuable to complement solution information
concerning aminoguanidine reactivity, since the gas-phase
data obtained here correlates well with literature solution data
for the reaction systems investigated.
experimental model of protein glycation”, Free Rad. Biol.
Med. 21, 801 (1996). doi: 10.1016/0891-5849(96)00183-9
11. Y. Nohara, T. Usui, T. Kinoshita and M. Watanabe,
“Generation of superoxide anions during the reaction
of guanidino compounds with methylglyoxal”, Chem.
Pharm. Bull. 50, 179 (2002). doi: 10.1248/cpb.50.179
12. D.L. Price, P.M. Rhett, S.R. Thorpe and J.W. Baynes,
“Chelating activity of advanced glycation end-products
inhibitors”, J. Biol. Chem. 276, 48967 (2001). doi: 10.1074/
jbc.M108196200
13. L.S. Santos, L. Knaack and J.O. Metzger, “Investigation
of chemical reactions in solution using API-MS”, Int. J.
Mass Spectrom. 246, 84 (2005) (and references therein).
doi: 10.1016/j.ijms.2005.08.016
14. M.N. Eberlin, “Electrospray ionization mass spectrom-
etry: a major tool to investigate reaction mechanism in
both solution and the gas phase”, Eur. J. Mass Spectrom.
13, 19 (2007). doi: 10.1255/ejms.837
Acknowledgements
The authors gratefully acknowledge Dr Carlos Cordeiro
and Professor Ana Ponces Freire of the enzymology group
(CQB-FCUL, Portugal) for their inestimable contribution to this
work. Discussions with Professor Susana Santos (CQB-FCUL,
Portugal) are also acknowledged. One of the authors, M. A.
Saraiva, thanks Fundação para a Ciência e a Tecnologia (FCT)
for financial support.
15. M.A. Saraiva, C.M. Borges and M.H. Florêncio,
“Behaviour of 4-(-2-hydroxyethyl)-1-piperazineethane-
sulfonic acid under electrospray ionisation mass spec-
trometry conditions,” Eur. J. Mass Spectrom. 16, 199
(2010). doi: 10.1255/ejms.1061
References
1. M. Brownlee, H. Vlassara, A.Kooney, P. Ulrich and A.
Cerami, “Aminoguanidine prevents diabetes-induced
arterial wall protein cross-linking”, Science 232, 1629
(1986). doi: 10.1126/science.3487117
16. J.G. Erickson, “3-Amino-as-triazines”, J. Am. Chem. Soc.
74, 4706 (1952). doi: 10.1021/ja01138a506
2. I. Mišur and Z. Turk, “Substituted guanidine compounds
as inhibitors of nonenzymatic glycation in vitro”, Croat.
Chem. Acta. 74, 455 (2001).
3. S. Rahbar and J.L. Figarola, “Inhibitors and breakers of
advanced glycation endproducts (AGEs): a review”, Curr.
Med. Chem.—Imun. Endoc. Metab. Agents 2, 135 (2002).
doi: 10.2174/1568013023358889
4. V.M. Monnier, “Intervention against the Maillard reac-
tion in vivo”, Arch. Biochem. Biophys. 419, 1 (2003). doi:
10.1016/j.abb.2003.08.014
17. T.W. Lo, T. Selwood and P.J. Thornalley, “Reaction of
methylglyoxal with aminoguanidine under physiological
conditions and prevention of methylglyoxal binding to
plasma proteins”, Biochem. Pharmacol. 48, 1865 (1994).
doi: 10.1016/0006-2952(94)90584-3
18. T.L. Constantopoulos, G.S. Jackson and C.G. Enke,
“Effects of salt concentration on analyte response using
electrospray ionization mass spectrometry”, J. Am.
Soc. Mass Spectrom. 10, 625 (1999). doi: 10.1016/S1044-
0305(99)00031-8
5. P.J. Thornalley, “Use of aminoguanidine (Pimagedine)
to prevent the formation of advanced glycation end-
product”, Arch. Biochem. Biophys. 419, 31 (2003). doi:
10.1016/j.abb.2003.08.013
6. P.V. Bharatam, P. Iqbal, A. Maldi and R. Tiwari, “Electron
delocalization in aminoguanidine: a computational
study”, J. Phys. Chem. A 108, 10509 (2004). doi: 10.1021/
jp049366e
19. M.A. Saraiva, C.M. Borges and M.H. Florêncio,
“Reactions of a modified lysine with aldehydic and
ketonic dicarbonyl compounds: an electrospray mass
spectrometry structure activity study”, J. Mass Spectrom.
41, 216 (2006). doi: 10.1002/jms.980
20. M.A. Saraiva, C.M. Borge and M.H. Florêncio, “Non-
enzymatic model glycation reactions—a comprehensive
study of the reactivity of a modified arginine with aldehy-
dic and diketonic dicarbonyl compounds by electrospray
mass spectrometry”, J. Mass Spectrom. 41, 755 (2006).
doi: 10.1002/jms.1031
7. E. Lieber and G.B.L. Smith, “The chemistry of amino-
guanidine and related substances”, Chem. Rev. 25, 213
(1939). doi: 10.1021/cr60081a003
8. P. Ou and S.P. Wolf, “Aminoguanidine: a drug proposed
for prophylaxis in diabetes inhibits catalase and gener-
ates hydrogen peroxyde in vitro”, Biochem. Pharmacol.
46, 1139 (1993). doi: 10.1016/0006-2952(93)90461-5
9. A. Philis-Tsimikis, S. Parthasarathy, S. Picard, W.
Palinski and J. Witztum, “Aminoguanidine has both pro-
oxidant and antioxidant activity toward LDL”, Arterioscler.
Thromb. Vasc. Biol. 15, 367 (1995). doi: 10.1161/01.
ATV.15.3.367
21. M.A. Saraiva, C.M. Borge and M.H. Florêncio, “Towards
the control and inhibition of glycation—the role of the
guanidine reaction center with aldehydic and dike-
tonic dicarbonyls. A mass spectrometry study. J. Mass
Spectrom. 41, 1346 (2006). doi: 10.1002/jms.1109
22. J. Sunner, G. Nicol and P. Kebarle, “Factors determining
relative sensitivity of analytes in positive mode atmos-
pheric pressure ionization mass spectrometry”, Anal.
Chem. 60, 1300 (1988). doi: 10.1021/ac00164a012
23. P.J. Thornalley, A. Yurek-George and O.K. Argirov,
“Kinetics and mechanism of the reaction of
10. J.T. Skamarauskas, A.G. McKay and J.V. Hunt,
“Aminoguanidine and its pro-oxidant effects on an

M.A. Saraiva, C.M. Borges and M.H. Florêncio, Eur. J. Mass Spectrom. 18, 385–397 (2012)
397
aminoguanidine with the a-oxoaldehydes glyoxal, meth-
ylglyoxal, and 3-deoxyglucosone under physiological
conditions”, Biochem. Pharmacol. 60, 55 (2000). doi:
10.1016/S0006-2952(00)00287-2
Guillaumet, “Highly diastereoselective synthesis of C(6)-
functionalized dihydroimidazotriazines”, Org. Lett. 5,
4595 (2003). doi: 10.1021/ol035743e
26. H. Neunhoeffer, in Comprehensive heterocyclic chemistry
II, Chapter 6.11, Ed by A.R. Katrizky, C.W. Rees and E.F.V.
Scriven. Pergamon Press, Oxford, UK, pp. 507–573
(1996).
27. P.V. Bernhardt and E.J. Hayes, “Aminotriazines as lock-
ing fragments in macrocyclic synthesis”, Inorg. Chem. 37,
4214 (1998). doi: 10.1021/ic980243g
24. J Limanto, R.A. Desmond, D.R. Gauthier, Jr, P.N.
Devine, R.A. Reamer and R.P. Volante, “A regioselective
approach to 5-substituted-3-amino-1,2,4-triazines”, Org.
Lett. 5, 2271 (2003). doi: 10.1021/ol034602+
25. E. Garnier, J. Guillard, E. Pasquinet, F. Suzenet,
D. Poullain, C. Jarry, J.-M. Léger, B. Lebret and G.