Oxygen-dependent fragmentation reactions during the degradation of 1-deoxy-d-erythro-hexo-2,3-diulose

Article abstract of DOI:10.1021/jf100140h

With this work, we report on further insights into the chemistry of 1-deoxy-d-erythro-hexo-2,3-diulose (1-deoxyglucosone, 1-DG). This α-dicarbonyl plays an important role as a highly reactive intermediate in the Maillard chemistry of hexoses. Degradation of 1-DG in the presence of the amino acid l-alanine led to the formation of several products. Lactic acid and glyceric acid were found to be major degradation products. Their formation was dependent on the presence of oxygen. Therefore, a mechanism is postulated based on oxidation leading to a tricarbonyl intermediate. Carbonyl cleavage of this structure should then give rise to carboxylic acids. This mechanism was supported by the isotope distribution observed during degradation of different ¹³C-labeled d-glucose isotopomers. Furthermore, we identified 3,5-dihydroxy-6-methyl-2,3-dihydro-4H-pyran-4-one (γ-pyranone) to be capable of rehydration forming 1-DG to a minor extent and therefore leading to the same degradation products. The formation of carboxylic acids from γ-pyranone was also dependent on the presence of oxygen in agreement with the postulated oxidative fragmentation. Finally, we investigated the formation of aldehydes expected as retro-aldol products formed within the degradation of 1-DG. Results seemed to rule out this reaction as an important degradation pathway under the conditions investigated herein.

Full text of DOI:10.1021/jf100140h

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
J. Agric. Food Chem. 2010, 58, 5685–5691 5685
DOI:10.1021/jf100140h
Oxygen-Dependent Fragmentation Reactions during the
Degradation of 1-Deoxy- -erythro-hexo-2,3-diulose
D
M
ICHAEL
V
OIGT, MAREEN
S
MUDA, CHRISTOPH
P
FAHLER
,
AND
M
ARCUS A. GLOMB
*
Institute of Chemistry, Food Chemistry, Martin-Luther-University Halle-Wittenberg,
Kurt-Mothes-Strasse 2, 06120 Halle/Saale, Germany
With this work, we report on further insights into the chemistry of 1-deoxy-
D-erythro-hexo-2,3-diulose
(1-deoxyglucosone, 1-DG). This R-dicarbonyl plays an important role as a highly reactive inter-
mediate in the Maillard chemistry of hexoses. Degradation of 1-DG in the presence of the amino
acid L-alanine led to the formation of several products. Lactic acid and glyceric acid were found to be
major degradation products. Their formation was dependent on the presence of oxygen. Therefore,
a mechanism is postulated based on oxidation leading to a tricarbonyl intermediate. Carbonyl
cleavage of this structure should then give rise to carboxylic acids. This mechanism was supported
by the isotope distribution observed during degradation of different 13C-labeled
D-glucose isotopo-
mers. Furthermore, we identified 3,5-dihydroxy-6-methyl-2,3-dihydro-4H-pyran-4-one (γ-pyranone)
to be capable of rehydration forming 1-DG to a minor extent and therefore leading to the same
degradation products. The formation of carboxylic acids from γ-pyranone was also dependent on the
presence of oxygen in agreement with the postulated oxidative fragmentation. Finally, we investi-
gated the formation of aldehydes expected as retro-aldol products formed within the degradation of
1-DG. Results seemed to rule out this reaction as an important degradation pathway under the
conditions investigated herein.
KEYWORDS: Maillard reaction; 1-deoxyhexo-2,3-diulose; 1-deoxyglucosone; 1-DG; r-dicarbonyl com-
pounds; dicarbonyl cleavage; oxidative degradation; tricarbonyl; carboxylic acids
INTRODUCTION
of 1-DG and its degradation products in detail. We recently
reported on the degradation of 1-DG in aqueous model systems
The term Maillard reaction is used today as a synonym for a
complicated network of different reaction pathways. In the
historic view, these reactions are initiated by reaction of reducing
sugars and amino acids or other compounds bearing an amino
group. Maillard reactions provide positive and negative aspects,
for example, the formation of flavor and color of processed foods.
Highly reactive intermediates, with an R-dicarbonyl moiety, are
able to react with amino acids and thereby influence the nutri-
tional value of processed foods (review in ref 1). Furthermore,
with the discovery of glycated hemoglobin (2, 3), it became
obvious that Maillard reactions also occur in vivo; thus, the
formation of so-called advanced glycation end products (AGE) is
discussed in the context of aging and complications like diabetes,
Alzheimer’s, or arteriosclerosis (4, 5). Therefore, Maillard reac-
tions and, in particular, the chemistry of R-dicarbonyl com-
pounds are issues of vital interest in food chemistry as well as in
in the presence of the amino acid L-alanine. Carboxylic acids and
dicarbonyl as well as hydroxycarbonyl compounds were among
the degradation products verified. Their formation was explained
nonoxidatively via hydrolytic β-dicarbonyl cleavage (8). In this
work, we extend the range of products identified, as well as
the knowledge on reaction pathways important to 1-DG degra-
dation.
MATERIALS AND METHODS
Materials. The following chemicals of analytical grade were commer-
cially available: N,O-bis-(trimethylsilyl)acetamide with 5% trimethyl-
chlorosilane,
L-alanine, L(þ)-erythrulose, o-phenylenediamine (OPD), hep-
tafluorobutyric acid,
L
-(þ)-lactic acid, 2,3,5-triphenyltetrazolium chloride
(Fluka/Sigma-Aldrich, Seelze, Germany), dipotassium hydrogen phos-
phate trihydrate, potassium dihydrogen phosphate (Merck, Darmstadt,
Germany), calcium sulfate, diethyl ether, ethyl acetate (EtOAC) (Roth,
Karlsruhe, Germany), diethylenetriaminepentaacetic acid (Sigma-Aldrich,
medical science. The R-dicarbonyl compound 1-deoxy-D-erythro-
hexo-2,3-diulose (1-deoxyglucosone, 1-DG) (1) is the key inter-
mediate in the Maillard-induced degradation of hexoses. Because
of its high reactivity, it was for a long time only accessible via
indirect analytical means like trapping reactions (6). Successful
synthesis of 1 (7) enabled us to investigate the chemical properties
Taufkirchen, Germany), D-glyceric acid calcium salt dihydrate (Aldrich/
Sigma-Aldrich, Steinheim, Germany), and Dowex 50 W ꢀ 8 (Hþ-form,
50-100 mesh) (Serva/Boehringer, Ingelheim-Heidelberg, Germany).
1-Deoxy-4,5-O-isopropylidene-D-erythro-hexo-2,3-diulose and 1-DG
were synthesized as described in ref 7.
Degradation of 1-DG. Degradation of 1-DG was carried out under
aerated and deaerated conditions according to ref8. Incubation solutions
(1-DG/
L-alanine, 42 mM in 0.1 M phosphate buffer, pH 7.4) were heated
*To whom correspondence should be addressed. Fax: þþ049-345-
5527341. E-mail marcus.glomb@chemie.uni-halle.de.
(37 and 50 °C), and samples were taken over time and analyzed by GC
© 2010 American Chemical Society
Published on Web 04/08/2010
pubs.acs.org/JAFC

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
5686 J. Agric. Food Chem., Vol. 58, No. 9, 2010
Voigt et al.
after derivatization. Values were expressed as means of at least three
independent determinations.
combined organic extracts were dried over calcium sulfate and evaporated
to dryness in vacuo. The residue was purified using a LiChroprep RP-18
column with water/methanol (90/10, v/v) as the eluent. Fractions with
target material (TLC, Rf 0.12, same solvent system) were combined,
Degradation of 3,5-Dihydroxy-6-methyl-2,3-dihydro-4H-pyran-
4-one (γ-Pyranone). γ-Pyranone was synthesized according to ref 9.
In brief, a solution of glucose and piperidine in ethanol was refluxed
under introduction of argon. After the addition of acetic acid and further
heating, the concentrated reaction solution was diluted with water, and
γ-pyranone was extracted with EtOAc. The crude product was purified
using column chromatography (silica gel) and high-vacuum distillation.
Finally, the product was recrystallized from hexane.
1
yielding the product as a colorless solid (199 mg, 1.12 mmol, 53%). H
NMR (200 MHz in CH3OD): δ 3.88 [dd, 1H, 3J = 6.7 Hz, 2J = 11.4 Hz,
-CHaHb(OH)], 4.05 [dd, 1H, 3J = 4.0 Hz, 2J = 11.2 Hz, -CHaHb-
(OH)], 5.02 [dd, 1H, 3J = 4.0 Hz, 3J = 6.5 Hz, -CH(OH)], 7.23 (m, 2H),
7.58 (m, 2H) ppm. HR-MS (after silylation) verified a molecular mass of
m/z 322.1537 (found) [m/z 322.1533 calculated for C15H26O2N2Si2 (Mþ•)].
GC-EI-MS. Samples were analyzed on a Thermo Finnigan Trace GC
Ultra coupled to a Thermo Finnigan Trace DSQ (both Thermo Fisher
Scientific GmbH, Bremen, Germany) using the parameters and tempera-
ture program given in ref 8.
Degradationof13C-LabeledGlucose. Different13C-labeled
isotopomers (labeled positions 1, 2, 3, 5, and 6) (12.6 μmol, respectively)
were dissolved together with -alanine (12.6 μmol) in phosphate buffer
D-glucose
L
(0.1 M, pH 7.4, 0.3 mL). The incubation solution was reacted at 50 °C for
7 days under aerobic conditions. After the solution was cooled to room
temperature, an aliquot (50 μL) was measured by GC-LCI-MS after
silylation for glyceric acid and by liquid chromatography-mass spectro-
metry (LC-MS2) directly for lactic acid.
Derivatization Reactions. Trimethylsilyl Derivatives. Trimethyl-
silyl derivatives were obtained by adopting the method described in
ref 8. Data for silylated compounds analyzed by gas chromatography-
mass spectrometry (GC-MS) showed standard deviations <10 mmol/mol
1-DG, resulting in coefficients of variation <5%.
GC-LCI-MS. GC-MS with liquid chemical ionization was performed
on the instrument described above for glyceric acid isotope label experi-
ments. Methanol was used as a reactant gas. Mass spectra were obtained at
70 eV (source, 190 °C; emission current, 80 μA) in full scan mode (mass
range m/z 50-650).
LC-MS2. LC-MS2 was applied for lactic acid isotope label experi-
ments in negative-ion mode. Mass transitions for multiple reaction moni-
toring (MRM) mode were set according to ref 12. The optimized para-
meters for mass spectrometry were as follows: m/z 89 f 43 (DP, -50.00;
CE, -19.20; CXP, 0.00), m/z 89 f 45 (DP, -48.90; CE, -15.20; CXP,
0.00), m/z 90 f 43 and m/z 90 f 44 (DP, -50.00; CE, -19.20; CXP, 0.00),
m/z 90 f 45 and m/z 90 f 46 (DP, -48.90; CE, -15.20; CXP, 0.00); the
dwell time was always 75.00 ms. H2CdCH-O- with m/z 43 (unlabeled
species) or m/z 44 (labeled species) represents -HCOH-CH3 of the lactic
acid molecule. -COOH with m/z 45 (unlabeled species) or m/z 46 (labeled
species) represents -COOH of the lactic acid molecule. A Jasco PU-2080
Plus quaternary gradient pump with degasser and a Jasco AS-2057 Plus
autosampler (Jasco, Gross-Umstadt, Germany) were used. Chromato-
graphic separations were performed on a stainless steel column (Eurospher
100-5 C18, 250 mm ꢀ 4.0 mm, Knauer, Berlin, Germany) using a flow
rate of 1 mL/min. The mobile phase used consisted of water (solvent A)
and MeOH/water [7:3 (v/v), solvent B]. To both solvents (A and B),
0.8 mL/L formic acid was added. Samples were injected at 100% A (held
5 min), and the gradient then changed to 100% B in 5 min (held 15 min)
and then changed to 100% A in 5 min (held 15 min). Lactic acid eluted at
tR = 4.0 min. The mass analyses were performed using an Applied
Biosystems API 4000 quadrupole instrument (Applied Biosystems, Foster
City, CA) equipped with an API source using an electrospray ionization
(ESI) interphase. The LC system was connected directly to the probe of the
mass spectrometer. Nitrogen was used as the sheath and auxiliary gas.
HPLC-UV. A Jasco PU-2089 Plus quaternary gradient pump with
degasser was used combined with a Jasco AS-2055 Plus autosampler. Elution
of benzimidazoles (glyceric aldehyde 11 at 27.8 min, glycolaldehyde 12 at
30.7 min, and acetaldehyde 13 at 38.9 min) and 1-DG-quinoxaline 10 (23.5
min) was monitored by a Jasco UV-2075 Plus UV detector (all Jasco, Gross-
Umstadt, Germany). Chromatographic separation was carried out on a
stainless steel column (Eurospher 100-5 C18, 250 mm ꢀ 4.6 mm) by Knauer
(Berlin, Germany) using the system described in our previous paper (8).
Accurate Mass Determination (HR-MS). The high-resolution
positive and negative ion ESI mass spectra (HR-MS) were obtained from
a Bruker Apex III Fourier transform ion cyclotron resonance (FT-ICR)
mass spectrometer (Bruker, Daltonics, Billerica, United States) equipped
with an Infinity cell, a 7.0 T superconducting magnet (Bruker, Karlsruhe,
Germany), a radio frequency-only hexapole ion guide, and an external
electrospray ion source (APOLLO, Agilent, off-axis spray). Nitrogen was
used as the drying gas at 150 °C. The samples were dissolved in methanol,
and the solutions were introduced continuously via a syringe pump at a
flow of 120 μL/h. The data were acquired with 256k data points and zero
filled to 1024k by averaging 32 scans.
Benzimidazole Derivatives. Samples were spiked with OPD (4.95 μmol)
dissolved in water (30 μL) and kept for 5 h at room temperature prior to
injection into the high-performance liquid chromatography-ultraviolet
(HPLC-UV) system. Benzimidazoles were monitored at λΜ = 272 nm.
Quantification was carried out by comparison of peak areas obtained at
272 nm with those of standard solutions containing known amounts of
pure authentic benzimidazoles (data obtained with HPLC-UV showed
standard deviations <1.5 mmol/mol 1-DG, resulting in coefficients of
variation <5%). Reference compounds were synthesized as follows.
2-Methyl-1H-benzimidazole. OPD (1.1 g, 10.2 mmol) was dis-
solved in phosphate buffer (20 mL, 0.1 M, pH 7.4). Acetaldehyde (381 μL,
6.8 mmol) was added, and the mixture was maintained for 3 h at room
temperature. After dilution with water (20 mL), the solution was extracted
with three volumes of 50 mL of EtOAc. The combined organic extracts
were dried over calcium sulfate and evaporated to dryness in vacuo. The
crude product was chromatographed on a silica gel column (3 cm ꢀ 15 cm)
with EtOAc/hexane (2/98, v/v) as an eluent. Fractions with target material
[thin-layer chromatography (TLC), Rf 0.15, same solvent system]
were combined, and the organic solvent was removed. The residue was
further purified using a LiChroprep RP-18 column with water/methanol
(70/30, v/v) as the eluent. Fractions containing the target material were
combined, yielding the product as a colorless solid (149 mg, 1.13 mmol,
17%). 1H NMR (200 MHz in CD3OD): δ 2.57 (s, -CH3), 7.20 (m, 2H),
7.50 (m, 2H) ppm. 13C NMR (200 MHz in CD3OD): δ 14.5, 115.4, 123.7,
139.4 153.2 ppm. High-resolution mass spectrometry (HR-MS) verified a
molecular mass of m/z 132.0684 (found) [m/z 132.0687 calculated for
C8H8N2 (Mþ•)]. NMR data were in line with refs 10 and 11.
1H-Benzimidazol-2-ylmethanol. OPD (866 mg, 8 mmol) was
dissolved in phosphate buffer (20 mL, 0.1 M, pH 7.4). Glycolaldehyde
dimer (303 mg, 2.5 mmol) was added, and the mixture was maintained for
3 h at room temperature. After dilution with water (20 mL), the solution
was extracted with three volumes of 50 mL of EtOAc. The combined
organic extracts were dried over calcium sulfate and evaporated to dryness
in vacuo. The crude product was chromatographed on a silica gel column
(3 cm ꢀ 15cm) with EtOAc/hexane(2/98, v/v) as the eluent. Fractions with
target material (TLC, Rf 0.12, same solvent system) were combined, and
the organic solvent was removed. The residue was further purified using a
LiChroprep RP-18 column with water/methanol (90/10, v/v) as the eluent.
Fractions containing the target were combined, yielding the product as a
colorless solid (142 mg, 0.96 mmol, 19%). 1H NMR (200 MHz in
CH3OD): δ 4.88 (s, 2H, -CH2OH), 7.22 (m, 2H), 7.56 (m, 2H) ppm.
HR-MS (after silylation) verified a molecular mass of m/z 220.1032
(found) [m/z 220.1032 calculated for C11H16ON2Si (Mþ•)].
Nuclear Magnetic Resonance Spectroscopy (NMR). 1H and 13C
experiments were performed on a Bruker AC-200 (Bruker, Rheinstetten,
Germany).
1-(1H-Benzimidazol-2-yl)ethane-1,2-diol. OPD (452 mg, 4.2
mmol) was dissolved in phosphate buffer (20 mL, 0.1 M, pH 7.4). Glyceric
aldehyde (189 mg, 2.1 mmol) was added, and the mixture was maintained
for 16 h at room temperature. After dilution with water (20 mL), the
solution was extracted with three volumes of 50 mL of EtOAc. The
RESULTS
Degradation of 1-DG. Recently, we described silylation of a
1-DG/L-alanine incubation (42 mM in phosphate buffer 0.1 M,
pH 7.4) leading to several signals in GC-MS (8). Major signals

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Article
J. Agric. Food Chem., Vol. 58, No. 9, 2010 5687
Table 1. Degradation of Different 13C-D-Glucose Isotopomers Revealed
the Formation of Discrete Lactic Acid Species from Different Parts of
the Original Carbon Backbone
Figure 1. Time-course formation of lactic acid (b, 37 °C; 9, 50 °C) in
incubations of 1-DG and L-alanine (42 mM, pH 7.4, respectively) under
aeration (full symbols) and under deaeration (open symbols).
were assigned to glyceric acid 3, erythrulose 4, γ-pyranone 5, and
4-hydroxy-2-(hydroxymethyl)-5-methylfuran-3(2H)-one 6 (furan-
3-one). Another product formed from 1 was lactic acid 2. Figure 1
shows the formation of lactic acid over a period of 48 h at two
different temperatures (37 and 50 °C) under aerated and deaerated
conditions. The formation of lactic acid was significantly preferred
under aerated conditions. Up to 166 mmol lactic acid per mol
1-DG was formed at 37 °C after 48 h (aerated conditions), while
incubation under deaerated conditions yielded only 26 mmol/mol
1-DG. The formation of lactic acid proceeded at a higher rate
within the first 2 h, that is, about half of the total amount
(87 mmol/mol 1-DG) was already formed. Later on, the concen-
tration of lactic acid increased at a lower rate but still nearly linear
over the whole time monitored. As reported before, the half-life of
1-DG was about 0.5 h under the given conditions (7). Incubations
at 50 °C resulted in significantly lower quantities of lactic acid as
compared to incubations at 37 °C. For example, only 118 mmol/
mol 1-DG was obtained after 48 h under aeration. Still, the diffe-
rence to deaeration was obvious (28 mmol/mol 1-DG). Interest-
ingly, lactic acid followed almost the same pattern in the rate of
formation and in concentration as glyceric acid. The largest diffe-
rences between these two compounds were observed at 37 °C under
aerated conditions. Here, the formation of glyceric acid was
favored (266 vs 166 mmol/mol 1-DG) under aerated conditions.
As with lactic acid, there was a significant increase for glyceric acid
under aeration and a significant decrease for comparable incuba-
tions conducted at 50 °C.
Figure 2. Formation of 3,5-dihydroxy-6-methyl-2,3-dihydro-4H-pyran-
4-one (γ-pyranone) 5 (1-DG/L-alanine, 42 mM in 0.1 M phosphate buffer,
pH 7.4; b, 37 °C; 9, 50 °C) under aeration (full symbols) and under
deaeration (open symbols).
with the -COOH group strictly located at C-3. Thirty-six percent
resulted from C-3 to C-5 with the -COOH group strictly at
C-5. The remaining 16% was verified from C-4 to C-6 with the
-COOH group assigned to C-4 and C-6 in equal parts.
Another product identified in the degradation of 1 was
γ-pyranone 5. Figure 2 shows the time-course formation of 5
under aerated and deaerated conditions (37 and 50 °C). Maximum
yield resulted after 2-3 h (173 mmol/mol 1-DG, 37 °C under de-
aeration) that decreased in later samples. Formation was preferred
under deaeration; for example, 146 mmol/mol 1-DG was detected
after 12 h (37 °C) under deaerated conditions vs 67 mmol/mol
1-DG under aerated conditions. Yields obtained at 50 °C showed
fewer differences (114 vs 103 mmol/mol 1-DG after 12 h).
We also expected the formation of aldehydes during the degra-
dation of 1. Therefore, samples taken at various time points were
reincubated with OPD and analyzed by HPLC-UV. In general,
we observed a rather fast formation of aldehydes within the first
5 h. As indicated in Figure 3 for acetaldehyde, concentrations
did almost not change at later incubation times. The highest
yields were obtained for acetaldehyde and glycolaldehyde at
50 °C under aerated conditions (30 and 29 mmol/mol 1-DG at
12 h, respectively), while glyceric aldehyde was only formed up to
Because of the lack of stable isotope-labeled 1, we investigated
the degradation of 13C-labeled glucose in the presence of
L-alanine to gain more information on the occurring reaction
pathways leading to the formation of glyceric acid and lactic acid.
In GC-LCI-MS after silylation, m/z 323 related to unlabeled
glyceric acid (M þ 1), while m/z 324 indicated label incorporation
(M þ 2). Glyceric acid always incorporated the label from C-5
and C-6 but not from C-1 to C-3 of glucose. This means that
glyceric acid is formed exclusively from the lower three carbon
atoms of glucose, confirming data from Davidek et al. (13). In
contrast, results obtained for lactic acid from LC-MS2 analyses
were not as clear-cut and are depicted in Table 1. Selected mass
transitions (12) not only allowed us to monitor label incorpora-
tion but also allowed us to locate the exact position of the label in
the -COOH versus the -CHOH-CH3 part of the lactic acid
molecule. The data unequivocally showed that 48% of lactic acid
formed during glucose degradation originated from C-1 to C-3

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
5688 J. Agric. Food Chem., Vol. 58, No. 9, 2010
Voigt et al.
Figure 4. Total ion chromatogram of a γ-pyranone/
mixture(42mM in0.1Mphosphate buffer, pH7.4;after 12hat 50°C under
deaerated conditions). Signals are trimethylsilyl derivatives of lactic acid 2,
L
-alanine incubation
Figure 3. Formation of acetaldehyde (1-DG/L-alanine, 42 mM in 0.1 M
phosphate buffer, pH 7.4).
L
-alanine 8, phosphoric acid 7, glyceric acid 3, an unknown compound 9,
γ-pyranone 5, erythrulose 4, and 4-hydroxy-2-(hydroxymethyl)-5-methyl-
furan-3(2H)-one 6.
13 mmol/mol 1-DG. For all three aldehydes, yields responded
significantly to aeration. We also used this trapping reaction
to investigate a possible Strecker degradation of the applied
R-alanine. Incubations in the presence of β-alanine yielded 50%
of the amount of acetaldehyde obtained in upper incubations,
while concentrations of glycolaldehyde and glyceric aldehyde did
not change. The postincubation trapping conditions were opti-
mized for coinciding R-dicarbonyl structures. Experiments for the
conversion ofacetaldehyde, glycolaldehyde, and glyceric aldehyde
monitored by authentic benzimidazols yielded only recoveries of
27% for acetaldehyde, 38% for glycolaldehyde, and 76% for
glyceric aldehyde. These relatively poor yields could not be signi-
ficantly improved by changing the reaction conditions. However,
stringent reaction conditions led to an acceptable coefficient of
variation of smaller than 5% and also avoided artifact forma-
tion (8, 14, 24). Both parameters, R-dicarbonyls and aldehydes,
can therefore be quantified by the same trapping reaction.
chemistry of hexoses. The formation of glyceric acid from a
glucose/glycine model system was recently reported by Davidek
and co-workers (13). In an associated work (15), these authors
described the formation of large amounts of acetic acid in the
degradation of 1 under cooking conditions. This result was
explained by a hydrolytic β-dicarbonyl cleavage of the isomeric
1-deoxyhexo-2,4-diulose induced by nucleophilic attack of a
hydroxyl anion at the C-2 carbonyl function. This mechanism
is also capable of explaining the formation of glyceric acid
(nucleophilic attack at the C-4 carbonyl function). The corres-
ponding counterpart in this mechanism is 1-hydroxypropan-
2-one (acetol), which we established in our previous work (8).
However, the formation of glyceric acid as well as of lactic was
significantly favored under aerated conditions. This result cannot
be explained on the basis of a simple hydrolytic cleavage reaction.
Indeed, Davidek et al. also reported an additional oxidative
R-dicarbonyl cleavage under cooking conditions (13). This
mechanism is able to explain the formation of glyceric acid and
lactic acid based on the oxidation of the isomeric 1-deoxy-hexo-
3,4-diulose but was reported to occur only to negligible extent,
which is in contrast to the significant yield that we found herein.
In addition in our previous work, we were not able to verify
erythronic acid as the corresponding counterpart of acetic acid,
which is the major carboxylic acid during 1-DG degradation. A
possible scheme was suggested earlier by Beck and Ledl to explain
the oxygen-dependent formation of both acids. They reported on
the formation of glyceric and lactic acid in a glucose/glycine
model system and postulated formation as shown in Figure 5 (16).
In this mechanism, 1-DG, or more likely its corresponding R-oxo-
enediol reductone structure, is oxidized to give the indicated tri-
carbonyl compound. Metal-catalyzed oxidations of Maillard
intermediates have been published before (17). The cyclic form
of this structure was suggested to undergo a β-dicarbonyl cleav-
age forming the glyceric acid ester of lactic acid. Hydrolysis then
produces lactic acid and glyceric acid. They also suggested that
the same pathway could be entered from γ-pyranone. This is in
good agreement with our results on the degradation of γ-pyra-
none; that is, both acids were formed preferably under aerated
conditions. The postulated ester was identified by Beck and Ledl
during their investigation of the rather drastic oxidative treatment
of γ-pyranone with iodine. These reports are in line with our
Degradation of 3,5-Dihydroxy-6-methyl-2,3-dihydro-4H-pyran-
4-one (γ-Pyranone) (5). As described above, γ-pyranone was
found to be an important product formed from 1-DG. Given
that γ-pyranone itself represents a reactive Maillard intermediate,
one could expect that related degradation products should con-
tribute to the overall spectrum of products formed during
degradation of 1-DG. To test this hypothesis, γ-pyranone was
synthesized independently and thermally treated (37 and 50 °C,
respectively) in the presence of L-alanine under the same condi-
tions as described above for 1-DG. Figure 4 shows a total ion
chromatogram obtained after silylation of an incubation sample
(after 12 h at 50 °C). Interestingly, we monitored the same major
products that were already found in the degradation of 1-DG.
Signals were assigned to lactic acid 2, glyceric acid 3, erythrulose
4, and furan-3-one 6. Again, the formation of lactic acid and
glyceric acid was preferred under aerated conditions, and the
formation of glyceric acid exceeded the formation of lactic acid to
some extent; for example, 30 mmol/mol γ-pyranone of glyceric
acid vs 23 mmol/mol γ-pyranone lactic acid was formed after 12 h
at 37 °C (aerated). At 50 °C, degradation resulted in 66 mmol/mol
γ-pyranone of glyceric acid vs 61 mmol/mol γ-pyranone of lactic
acid (12 h/aerated conditions). Furthermore, we verified furan-
3-one 6 as a degradation product of 5.
DISCUSSION
Carboxylic acids represent stable degradation products formed
during degradation of 1-DG and therefore in the Maillard