Degradation of 1-deoxy-d-erythro-hexo-2,3-diulose in the presence of lysine leads to formation of carboxylic acid amides

Article abstract of DOI:10.1021/jf100334r

A novel species of amides formed from degradation of one of the most important key intermediates in Maillard hexose chemistry-1-deoxyhexo-2,3- diulose-was investigated. In 1-deoxyhexo-2,3-diulose/N^α-t-BOC- lysine reaction mixtures four amides, N^ε-acetyl lysine, N ^ε-formyl lysine, N^ε-lactoyl lysine and N ^ε-glycerinyl lysine, were identified and their structures verified by authentic reference standards. Amides and corresponding carboxylic acids (acetic acid, formic acid, lactic acid and glyceric acid) accumulated over time. Both N^ε-lysine amides and carboxylic acids were thus determined as stable Maillard end products. Results of model incubations suggested the synthesis of amides to be mechanistically closely related to the formation of their corresponding carboxylic acids by β-dicarbonyl cleavage. Due to the different chemical properties of all the compounds monitored, various analytical strategies had to be carried out (LC-MS², GC-MS, GC-FID, enzymatic determination).

Full text of DOI:10.1021/jf100334r

6458 J. Agric. Food Chem. 2010, 58, 6458–6464
DOI:10.1021/jf100334r
Degradation of 1-Deoxy- -erythro-hexo-2,3-diulose in
D
the Presence of Lysine Leads to Formation of Carboxylic
Acid Amides
MAREEN
SMUDA, MICHAEL
V
OIGT
,
AND
MARCUS A. GLOMB*
Institute of Chemistry, Food Chemistry, Martin-Luther-University Halle-Wittenberg,
Kurt-Mothes-Strasse 2, 06120 Halle/Saale, Germany
A novel species of amides formed from degradation of one of the most important key intermediates
in Maillard hexose chemistry-1-deoxyhexo-2,3-diulose-was investigated. In 1-deoxyhexo-2,3-
diulose/NR-t-BOC-lysine reaction mixtures four amides, Nε-acetyl lysine, Nε-formyl lysine, Nε-lactoyl
lysine and Nε-glycerinyl lysine, were identified and their structures verified by authentic reference
standards. Amides and corresponding carboxylic acids (acetic acid, formic acid, lactic acid and
glyceric acid) accumulated over time. Both Nε-lysine amides and carboxylic acids were thus
determined as stable Maillard end products. Results of model incubations suggested the synthesis
of amides to be mechanistically closely related to the formation of their corresponding carboxylic
acids by β-dicarbonyl cleavage. Due to the different chemical properties of all the compounds
monitored, various analytical strategies had to be carried out (LC-MS2, GC-MS, GC-FID,
enzymatic determination).
KEYWORDS: Maillard reaction; β-dicarbonyl cleavage; 1-deoxyhexo-2,3-diulose; 1-DG; Nε-lysine
amides; carboxylic acid amides; carboxylic acids
INTRODUCTION
in a similar way. Nagaraj et al. studied the formation of oxalate
monoalkylamide in the human lens in 1999 (11). In 2001
Glomb and Pfahler postulated the formation pathway of GOLA
(Nε-{2-[(5-amino-5-carboxypentyl)amino]-2-oxoethyl}lysine) and
GALA (Nε-glycoloyllysine) in glyoxal/lysine reaction mixtures.
With the detection of GOLA and GALA in brunescent lens
proteins the importance of these amide protein modifications with
respect to pathological processes in vivo was demonstrated (12).
Nε-Formyl lysine was already mentioned in the literature as a
glycation product of proteins, but a formation mechanism was not
given (13).
The present work extends the knowledge of amide formation in
Maillard reactions. The four amides Nε-acetyl lysine, Nε-formyl
lysine, Nε-lactoyl lysine, and Nε-glycerinyl lysine as well as the
corresponding carboxylic acids acetic acid, formic acid, lactic
acid, and glyceric acid were found to be formed in 1-DG/NR-t-
BOC-lysine reaction mixtures. Quantification of the compounds
was done by various analytical methods to gain further insights
into the degradation mechanisms of 1-DG.
During heating or storing of food, the Maillard reaction leads
to degradation of reducing sugars and amino acids, peptides or
proteins (1). In addition to the formation of color, taste and
flavor, the Maillard reaction results in a decline of the biological
value of proteins caused by reactions of the ε-amino group of
lysine (2). Inthe past decades the outstanding roleofR-dicarbonyl
compounds as intermediates of high reactivity in the complex
Maillard chemistry was manifested. Among these intermediates
1-deoxyhexo-2,3-diulose (1-DG) is ofmajor importanceinhexose
chemistry (3-5). Since Glomb and Pfahler succeeded in the
synthesis of 1-DG, it is possible to study carbohydrate Maillard
degradation mechanistically isolated on this intermediate (6).
Davidek et al. postulated a formation pathway of acetic acid
and glyceric acid that can be explained by a hydrolytic β-di-
carbonyl cleavage mechanism starting from 1-DG (7, 8). In our
study we now investigated a novel type of Maillard reaction
products: carboxylic acid amides formed by the ε-amino group of
lysine during degradation of 1-DG. In the literature, formation of
amides in Maillard chemistry is only scarcely described. In 1985
Hayase and co-workers investigated the formation of N-butyl-
acetamide and N-butylformamide in model reaction systems of
butylamine and the R-dicarbonyl structures diacetyl and glyoxal (9).
Several R-hydroxy acid amides such as lactic acid propylamide
have been identified in reaction mixtures of glucose and propyl-
amine (10), but it is still unknown if lysine side chains can react
MATERIALS AND METHODS
Materials. The following chemicals of analytical grade were commer-
cially available:N,O-bis(trimethylsilyl)acetamidewith 5% trimethylchloro-
silane, acetic acid, heptafluorobutyric acid (Fluka/Sigma-Aldrich, Seelze,
Germany), ninhydrin, dipotassium hydrogen phosphate trihydrate,
potassium dihydrogen phosphate (Merck, Darmstadt, Germany), decyl-
chloroformate, N-ethyl-N0-(3-dimethylaminopropyl)carbodiimide (EDC), 2,3-
O-isopropylidene-
D
-glyceric acid methyl ester,
D
-glyceric acid calcium salt,
NR-t-BOC-lysine, Nε-acetyl-
L
-lysine, Nε-formyl-
L-lysine (Sigma-Aldrich,
*To whom correspondence should be addressed. E-mail: marcus.
glomb@chemie.uni-halle.de. Fax: þþ049-345-5527341.
Steinheim, Germany), 1-hydroxybenzotriazole (HOBt), pyridine, thioanisole,
pubs.acs.org/JAFC
Published on Web 04/29/2010
© 2010 American Chemical Society

Article
J. Agric. Food Chem., Vol. 58, No. 10, 2010 6459
Nε-Glycerinyl-
L-lysine Trifluoroacetate (8). Deprotection and puri-
diethylenetriaminepentaacetic acid (Fluka/Sigma-Aldrich, Taufkirchen,
Germany), TFA (Carl Roth, Karlsruhe, Germany), Dowex 50W ꢀ 8 (Hþ-
form, 50-100 mesh) (Serva/Boehringer, Ingelheim-Heidelberg, Germany),
fication of 7 was done as described above for compound 4. 8 was obtained
as a colorless glassy foam in quantitative yield after lyophilization. 1H
NMR (400 MHz, D2O): δ [ppm] = 1.25-1.39 (m, 2H), 1.42-1.49 (m,
2H), 1.76-1.95 (m, 2H), 3.15 (t, 3J = 6.6 Hz, 2H), 3.62-3.71 (m, 2H), 3.93
(t, 3J = 6.0 Hz, 1H), 4.07-4.12 (m, 1H). 13C NMR (100 MHz, D2O): δ
[ppm] = 21.6, 28.0, 29.5, 38.6, 53.1, 63.4, 72.4, 172.3, 174.2. HR-MS: m/z
235.1288 (found); m/z 235.1286 (calculated for C9H19O5N2 [M þ H]þ).
Degradation of 1-Deoxyhexo-2,3-diulose. A solution of 1-deoxy-
€
L
-lactic acid (Alfa Aesar, Karlsruhe, Germany), and formic acid (Grussing
GmbH, Filsum, Germany). 1-Deoxy-4,5-O-isopropylidene- -erythro-
hexo-2,3-diulose and 1-deoxy-
-erythro-hexo-2,3-diulose (6), NR-(t-butoxy-
carbonyl)- -lactic
-lysine tert-butyl ester (1) (14), and OR-tetrahydropyranyl-
acid (2) (15) were synthesized according to the literature.
Nr-(t-Butoxycarbonyl)-Nε-(Or-(tetrahydropyranyl)-
D
D
L
L
L
-lactoyl)-L-lysine
tert-Butyl Ester (3). 2 (70 mg, 0.4 mmol) was dissolved in CH2Cl2 (2 mL)
under argon atmosphere at 0 °C, and HOBt (54 mg, 0.4 mmol) was added.
After 10 min a solution of EDC (80 μL, 0.45 mmol) in CH2Cl2 (1 mL) was
added dropwise. To the stirred solution, 1 (120 mg, 0.4 mmol) dissolved in
CH2Cl2 (1 mL) was added dropwise. The reaction mixture was warmed to
room temperature and stirred overnight. Then, the mixture was diluted
with CH2Cl2 (10 mL) and washed with 15 mL each of saturated NaHCO3
solution and brine. The organic layer was dried over MgSO4, and the
solvent was removed under reduced pressure. The crude product was purified
by column chromatography on silica gel 60 using EtOAc-hexane (1:1).
Fractions containing 3 (TLC: Rf 0.83 in EtOAc, ninhydrin detection) were
collectively concentrated in vacuo to afford compound 3 as a yellow viscous
oil (90 mg, 49%). 1H NMR (400 MHz, CDCl3): δ [ppm] = 1.42-1.82 (m,
12 H), 1.31 (d, 3J= 6.8 Hz, 3H), 1.38 (s, 9H), 1.39 (s, 9H), 3.13-3.24(m, 2H),
3.41-3.49 and 3.75-3.86 (m, 2H), 4.02-4.18 (m, 2H), 4.52-4.57 (m, 1H).
13C NMR (100 MHz, CDCl3): δ [ppm] = 17.9, 19.6/19.9, 20.4/22.6, 25.1/
25.2, 28.0, 28.4, 29.3/29.4, 31.0, 32.5, 38.6, 53.7/53.8, 62.8/64.0, 73.3/74.3, 79.5,
81.7, 98.4/98.9, 155.4, 171.7/171.8, 173.0/173.3. HR-MS: m/z 481.2884
(found); m/z 481.2886 (calculated for C23H42O7N2Na [M þ Na]þ).
4,5-O-isopropylidene-D-erythro-hexo-2,3-diulose (0.35 mmol) in water
(3 mL) was stirred with Dowex 50W ꢀ 8 (Hþ-form, 50-100 mesh, 4 mL)
for 4 h under argon atmosphere. The resin was filtered off and washed
with MeOH (3 ꢀ 2 mL). After evaporation of the combined solvents, the
residue was dissolved in phosphate buffer (0.1 M, pH 7.4, 4.2 mL). Aliquots
(150 μL) of this solution and a solution of NR-t-BOC-lysine (0.35 mmol) in
phosphate buffer (0.1 M, pH 7.4, 4.2 mL) were mixed in screw-cap vials
giving an incubation solution of 1-DG and NR-t-BOC-lysine (42 mM,
respectively). Incubation solutions were shaken at 37 °C, and samples were
taken over time.
Amides were analyzed by HPLC-MS2 after deprotection of the BOC
group. Lactic and glyceric acid were analyzed by GC-MS after silylation,
acetic acid by GC-FID as its decylchloroformate derivative and formic
acid by enzymatic determination. Each sample was prepared at least three
times.
Deaerated Incubations. Degradation of 1-deoxy-D-erythro-hexo-2,3-
diulose under deaerated conditions was carried out using phosphate buffer
with 1 mM diethylenetriaminepentaacetic acid. Buffer was degassed with
helium before samples were prepared; samples were deaerated with argon
before incubation.
Nε-
L-Lactoyl-L-lysine Trifluoroacetate (4). To 3 (66 mg, 0.14 mmol)
Deprotection Reaction of the Amides. To aliquots (100 μL) of the
incubation solutions 6 M HCl (100 μL) was added, and samples were kept
at room temperature for 30 min. Solutions were diluted on a scale of 1:100
with water prior to injection into the HPLC-MS2 system.
Control Experiment. Carboxylic acids (acetic acid, lactic acid, gly-
ceric acid and formic acid, 13 mM, respectively) were separately incubated
with NR-t-BOC-lysine (42 mM) in phosphate buffer (0.1 M, pH 7.4).
Incubation solutions were shaken at 37 °C, and samples were taken after
7 days. Further workup was performed analogous to the amide samples
prior to injection into the HPLC-MS2 system.
Derivatization Reactions for the Carboxylic Acids. Trimethylsilyl
Derivatives of Lactic Acid and Glyceric Acid. Adopting the method
described in ref 16, aliquots of the samples (50 μL) were dried in vacuo,
residues were dissolved in pyridine (50 μL), and N,O-bis(trimethylsilyl)-
acetamide with 5% trimethylchlorosilane (50 μL) was added. Samples
were kept 3 h at room temperature prior to injection into the GC-MS
system. Quantification was carried out by comparison of peak areas
obtained in the TIC with those of standard solutions containing known
amounts of the pure authentic reference compounds. Signals of target
compounds were standardized using the signal of silylated phosphoric acid
present in all samples. Data for silylated compounds obtained by GC-MS
showed standard deviations of <5 mmol/mol 1-DG, resulting in coeffi-
cients of variation <5%.
a solution of 25% thioanisole in TFA (8 mL) was added. The reaction
mixture was stirred at room temperature for 6 h under argon atmosphere
while the color of the solution changed from green to brown. The mixture
was concentrated in vacuo, and the residue was partitioned between 20 mL
each of diethyl ether and water. The organic layer was further extracted
with water (10 mL). The combined aqueous layers were concentrated
under reduced pressure to a final volume of 3 mL. The crude product was
purified by column chromatography (Lichroprep RP C18 material). The
column was pre-equilibrated with methanol (50 mL) followed by 0.1%
TFA in water (50 mL). The product was eluted with 0.1% TFA in water.
Fractions containing 4 (positive ninhydrin stain) were collectively lyophi-
lized to give a colorless glassy foam in quantitative yield. 1H NMR
3
(400 MHz, D2O): δ [ppm] = 1.23 (d, J = 6.8 Hz, 3H), 1.25-1.40 (m,
2H), 1.42-1.52 (m, 2H), 1.75-1.95 (m, 2H), 3.09-3.16 (m, 2H), 3.90-3.95
(m, 1H), 4.10 (q, 3J = 6.6 Hz, 1H). 13C NMR (100 MHz, D2O): δ = 19.8,
21.6, 28.0, 29.5, 38.4, 53.0, 67.8, 172.2, 177.4. HR-MS: m/z 219.1339
(found); m/z 219.1335 (calculated for C9H19O4N2 [M þ H]þ).
2,3-O-Isopropylidene-glyceric Acid (6). To a solution of KOH
(95 mg, 1.7 mmol) in 250 μL of water and 500 μL of absolute EtOH,
2,3-O-isopropylidene-D-glyceric acid methyl ester (5) (160 mg, 1 mmol)
was added. The reaction mixture was stirred 30 min at room temperature.
After adjusting the pH value to 4.3 by using HCl (1 M), solution was
extracted with EtOAc (5 ꢀ 3 mL), and the combined organic layers were
dried (MgSO4) and concentrated to give 62 mg (42%) of a colorless oil. 1H
NMR (400 MHz, CDCl3): δ [ppm] = 1.38 (s, 3H), 1.48 (s, 3H), 4.14 (dd,
3J = 5.0 and 2J = 8.9 Hz, 1H), 4.25 (dd, 3J = 7.5 and 2J = 8.7 Hz, 1H),
4.60 (dd, 3J = 5.0 and 3J = 7.5 Hz, 1H), 9.54 (br, 1H). 13C NMR (100
MHz, CDCl3): δ [ppm] = 25.3, 25.8, 67.2, 73.6, 111.8, 175.8.
Decylchloroformate Derivative of Acetic Acid. Adopting the method
described in ref17, samples (300 μL) were spiked with chlorosuccinic acid
(50 μg) dissolved in water as internal standard, pyridine (40 μL) and
decylchloroformate (50 μL) were added. The mixture was then sonicated
for 10 min. Decyl esters were extracted with hexane, and the organic layer
was analyzed by GC-FID. Quantitative results were obtained by internal
calibration using commercially available acetic acid. Data obtained by
GC-FID showed standard deviations of <10 mmol/mol 1-DG, result-
ing in coefficients of variation <3.5%.
Nr-(t-Butoxycarbonyl)-Nε-2,3-O-isopropylidene-glycerinyl-
L-lysine
tert-Butyl Ester (7). 6 was reacted in an equimolar ratio with HOBt
and 1 and with a 1.1-fold excess of EDC as described above for 3.
Purification was done by column chromatography (silica gel 60, EtOAc-
hexane 1:1). Fractions containing 7 (TLC: Rf 0.87 in EtOAc, ninhydrin
detection) were collectively concentrated in vacuo to afford 7 as a yellow
viscous oil (105 mg, 61%). 1H NMR (400 MHz, CDCl3): δ [ppm] =
1.24-1.44 (m, 2H), 1.33 (s, 3H), 1.38 (s, 9H), 1.40 (s, 9H), 1.42 (s, 3H),
1.46-1.60 (m, 3H), 1.70-1.80 (m, 1H), 3.21 (m, 2H), 4.01 (dd, 3J = 5.4
and 2J = 8.7 Hz, 1H), 4.03-4.13 (m, 1H) 4.21 (dd, 3J = 7.7 and 2J = 8.7
Hz, 1H), 4.40 (dd, 3J = 5.4 and 3J = 7.5 Hz, 1H). 13C NMR (100 MHz,
CDCl3): δ [ppm] = 22.5, 25.0, 26.1, 28.0, 28.3, 29.2, 32.5, 38.6, 53.7, 67.7,
75.0, 79.5, 81.7, 110.7, 155.2, 170.9, 171.6. HR-MS: m/z 453.2571 (found);
m/z 453.2570 (calculated for C21H38O7N2Na [M þ Na]þ).
Analytical HPLC-MS2. A Jasco PU-2080 Plus quaternary gradient
pump with degasser and Jasco AS-2057 Plus autosampler (Jasco, Gross-
Umstadt, Germany) were used. Chromatographic separations were per-
formed on a stainless steel column (VYDAC 218TP54, 250 ꢀ 4.6 mm,
RP18, 5 μm, Hesperia, CA) 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), 1.2 mL/L heptafluorobutyric acid
(HFBA) was added. Samples were injected at 2% B (held 25 min), gradient
then changed to 100% B in 5 min (held 10 min) and then changed to 2% B
in 5 min (held 15 min). Elution of amides (Nε-glycerinyl lysine at 7.0 min,

6460 J. Agric. Food Chem., Vol. 58, No. 10, 2010
Smuda et al.
Table 1. Mass Spectrometer Parameters for Amide Quantification (MRM Mode)
Q1 mass (amu)
Q3 mass (amu)
dwell time (ms)
DP
CE
CXP
Nε-acetyl lysine
Nε-formyl lysine
Nε-lactoyl lysine
Nε-glycerinyl lysine
189.20
175.10
219.20
235.30
126.10
112.10
156.20
84.20
75.00
75.00
75.00
75.00
30.00
25.00
32.00
48.00
18.00
20.00
20.00
37.00
10.00
13.00
8.00
6.00
Nε-formyl lysine at 9.3 min, Nε-acetyl lysine at 13.6 min, and Nε-lactoyl
lysine at 13.8 min) was monitored by mass detection. 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) interface. The LC
system was connected directly to the probe of the mass spectrometer.
Nitrogen was used as sheath and auxiliary gas. To measure the amides
in incubation solutions the MRM mode of HPLC-MS2 was used. The
optimized parameters for mass spectrometry are given in Table 1. Quanti-
fication was based on the standard addition method using solutions
containing known amounts of the pure authentic reference compounds.
Data for amides obtained by HPLC-MS2 showed standard deviations
of <0.07 mmol/mol 1-DG, resulting in coefficients of variation <5%.
GC-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). The GC column was a HP-5
(30 m ꢀ 0.32 mm, film thickness 0.25 μm; Agilent Technologies, Palo
Alto, CA); injector, 220 °C; split ratio, 1:30; transfer line, 250 °C. The oven
temperature program was as follows: 100 °C (0 min), 5 °C/min to 200 °C
(0 min), 10 °C/min to 270 °C (10min). Helium 5.0 was used as carrier gas in
constant flow mode (linear velocity 28 cm/s, flow 1 mL/min). Mass spectra
were obtained with EI at 70 eV (source: 210 °C) in full scan mode
(mass range m/z 50-650).
GC-FID. Samples were analyzed on a HP 6890N chromatograph
(Agilent Technologies, Palo Alto, CA) equipped with a flame ionization
detector. The column was a HP-5 (30 m ꢀ 0.32 mm, film thickness
0.25 μm; Agilent Technologies, PaloAlto, CA); injector, 250 °C; split ratio,
1:10; detector, 270 °C. Helium 4.6 was used as carrier gas in constant flow
mode (linear velocity 28 cm/s, flow 1.6 mL/min). The oven temperature
program was as follows: 50 °C (1 min), 4 °C/min to 120 °C (15 min),
40 °C/min to 220 °C (3 min), 30 °C/min to 260 °C (15 min).
Nuclear Magnetic Resonance Spectroscopy (NMR). NMR spec-
tra were recorded on a Varian VXR 400 spectrometer operating at 400
MHz for 1H and 100 MHz for 13C, respectively. Chemical shifts are given
relative to external SiMe4.
RESULTS
Syntheses of Authentic Amide References. Lactic acid amide (4,
Nε-lactoyl lysine) was synthesized in a typical amide formation
reaction (Figure 1 a), starting from lactic acid as the tetrahydro-
pyranyloxy derivative (2). 2 was activated with EDC and HOBt
and reacted with NR-(t-butoxycarbonyl)-
L-lysine tert-butyl ester
(1) to give the fully protected amide 3. In the last step all three
protective groups were completely removed by treatment with
thioanisole in TFA to result in the lactid acid amide 4.
In a similar procedure glyceric acid amide (8, Nε-glycerinyl
lysine) was synthesized (Figure 1 b). After activation of isopro-
pylidene protected glyceric acid (6) and reaction with 1 the
protected amide 7 was isolated. Finally 7 was deprotected under
acidic conditions to obtain 8.
Both target compounds 4 and 8 as well as the intermediates 3,
6, and 7 were verified by nuclear magnetic resonance experiments.
Furthermore the elemental composition of 3, 4, 7, and 8 was
confirmed by accurate mass determination.
Formation of Carboxylic Acids in 1-DG/Nr-t-BOC-lysine In-
cubations (pH 7.4, 37 °C). Degradation of 1-deoxyhexo-2,3-
diulose (1-DG) in the presence of amine led to the formation of
short chain carboxylic acids. Formation of formic, acetic, lactic
and glyceric acid was monitored over a period of seven days.
Results are shown in Figure 2 A. Within the first twelve hours a
rapid accumulation of all carboxylic acids was observed. The
further progress of the reaction was characterized by a marginal
increase of the acids until concentrations remained unchanged.
Formation of acetic acid led to the highest amounts after seven
days (310 mmol/mol 1-DG), followed by glyceric acid (88 mmol/
mol 1-DG), lactic acid (58 mmol/mol 1-DG) and formic acid
(27 mmol/mol 1-DG).
Formation of Carboxylic Acid Amides in 1-DG/Nr-t-BOC-
lysine Incubations (pH 7.4, 37 °C). Before incubation solutions
could be analyzed for the corresponding carboxylic acid amides,
samples were treated with acid to remove the BOC protection
group. Amides were found to be stable under these workup
conditions. A HPLC-MS2 method (MRM mode) was developed
to perform quantification of the amides in model incubation
systems. Collision induced dissociation (CID) experiments were
carried out to observe the fragmentation pattern of the different
amides. Table 2 demonstrates that all four amides underlie
analogous fragmentation steps. Fragmentation of Nε-lactoyl
lysine as an exemplary representative of analyzed amides can be
explained as follows: Based on QMþ ion of Nε-lactoyl lysine
(m/z = 219) loss of HCOOH gives the ion at m/z 173. Loss of
both HCOOH and ammonia at the R-amino group gives the ion
at m/z 156 with the highest abundance. Ions at m/z 130 and m/z 84
are fragments that are prominent in the mass spectra of all four
amides. m/z 84 presents the pyrrolinium ion. The ion at m/z 130
occurs after a cyclization step of Nε-lactoyl lysine to a six-membered
ring and loss of Nε-functionality. All these fragmentations are
similar to the fragmentation of ordinary lysine reported in refs 18
and 19. Accurate mass determination of synthesized Nε-lactoyl
lysine gave m/z 173.1284 for C8H17O2N2 (calculated m/z 173.1281)
and m/z 156.1019 for C12H14O2N (calculated m/z 156.1016) and
confirmed the results obtained by CID experiments.
Accurate Mass Determination (HR-MS). The high-resolution
positive 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, MA) equipped with an Infinity
cell, a 7.0 T superconducting magnet (Bruker, Karlsruhe, Germany), a
radio frequency (RF)-only hexapole ion guide, and an external electro-
spray ion source (Apollo; Agilent, off-axis spray). Nitrogen was used as
drying gas at 150 °C. The samples were dissolved in methanol, and the
solutions were introducedcontinuously via a syringe pump at a flowrate of
120 μL h-1. The data were acquired with 256 k data points and zero filled
to 1024 k by averaging 32 scans.
Electrospray Ionization Mass Spectrometry (ESI-MS2). The
mass analyses were performed using the instrument described above for
HPLC-MS2 analyses. The samples were dissolved in water, and the
solutions were introducedcontinuously via a syringe pump at a flowrate of
120 μL h-1. Spectra were obtained by using the following parameters: for
Nε-acetyl lysine (DP 50, CE 20, CXP 10), for Nε-formyl lysine (DP 50, CE
20, CXP 10), for Nε-lactoyl lysine (DP 55, CE 22, CXP 10), and for Nε-
glycerinyl lysine (DP 50, CE 24, CXP 10). In all cases 42 MCA scans were
carried out.
Enzymatic Determination of Formic Acid. Quantification of formic
acid in incubation solutions was carried out with an enzyme test kit
obtained from R-Biopharm AG (Darmstadt, Germany). Depending on
the amount of formic acid after different incubation times, varying
volumes of incubation solutions were directly used for the enzyme test
(100 to 500 μL). Coefficients of variation (<2.5%) were in line with the
test kit description.

Article
J. Agric. Food Chem., Vol. 58, No. 10, 2010 6461
Figure 1. Synthesis of Nε-lactoyl lysine and Nε-glycerinyl lysine.
Figure 2. Formation of carboxylic acids (A) and corresponding amides (B) in 1-DG/NR-t-BOC-lysine incubation mixtures (42 mM in phosphate buffer 0.1 M,
pH 7.4, at 37 °C under aerated conditions): acetic acid/Nε-acetyl lysine (2), glyceric acid/Nε-glycerinyl lysine (9), lactic acid/Nε-lactoyl lysine (b) and formic
acid/Nε-formyl lysine ([).
Table 2. Comparison of Fragmentation of the Different Amides in Positive Ion ESI-MS2-CID Experiments
fragmentation pattern, m/z (amu)
(M þ H)
(M þ H -HCOOH)
(M þ H - HCOOH - NH3)
(M þ H - H2N-CO-R)
pyrrolinium ion
Nε-acetyl lysine
Nε-formyl lysine
Nε-lactoyl lysine
Nε-glycerinyl lysine
189
175
219
235
143
129
173
189
126
112
156
172
130a
130b
130c
130d
84
84
84
84
a R = CH3. b R = H. c R = CH(OH)-CH3. d R = CH(OH)-CH2OH.
Formation of Nε-acetyl lysine, Nε-formyl lysine, Nε-lactoyl
lysine, and Nε-glycerinyl lysine was also monitored over a period
of seven days. The time course formation plot of the car-
boxylic acid amides is shown in Figure 2 B. In parallel with the
corresponding carboxylic acids there was a rapid increase of all
amideswithinthe first twelve hours. Alsoat later incubation times
concentrations reached a steady level. Unexpectedly, formation
of Nε-formyl lysine led to the highest amounts after seven days

6462 J. Agric. Food Chem., Vol. 58, No. 10, 2010
Smuda et al.
(1.48 mmol/mol 1-DG), followed by Nε-acetyl lysine (1.01 mmol/
mol 1-DG), Nε-lactoyl lysine (0.28 mmol/mol 1-DG), and Nε-
glycerinyl lysine (0.16 mmol/mol 1-DG).
in comparison to0.75 mmol/mol 1-DG and 0.26mmol/mol1-DG
under deaerated conditions, respectively.
Table 3 provides information about concentrations of car-
boxylic acid amides formed under aerated and under deae-
rated incubation conditions at 37 °C after 24 and 72 h. Under
aerated conditions we measured significant higher amounts of
Nε-glycerinyl lysine and of Nε-formyl lysine. 0.15 mmol/mol
1-DG of Nε-glycerinyl lysine and 1.47 mmol/mol 1-DG of
Nε-formyl lysine were formed after 72 h at 37 °C under aerated
conditions in contrast to 0.09 mmol/mol 1-DG and 1.08 mmol/mol
1-DG under deaerated conditions, respectively. Concentrations
of Nε-acetyl lysine and of Nε-lactoyl lysine under aerated versus
under deaerated conditions varied only slightly. 0.78 mmol/mol
1-DG of Nε-acetyl lysine and 0.25 mmol/mol 1-DG of Nε-lactoyl
lysine were obtained after 72 h at 37 °C under aerated conditions,
DISCUSSION
Carboxylic acids formed in Maillard reaction systems were
already established to be stable degradation products (7, 20).
Data obtained in the present work for the carboxylic acids
acetic acid, formic acid, lactic acid and glyceric acid confirmed
this finding. The corresponding carboxylic acid amides revealed
the same behavior. Over a period of seven days all amides
monitored accumulated to reach a plateau. Thus, these amides
can be regarded as stable degradation end products in Maillard
reactions too.
Davidek et al. described a hydrolytic β-dicarbonyl cleavage
mechanism of 1-deoxyhexo-2,4-diulose as the major formation
pathway for acetic acid as well as for glyceric acid (7), based on
studies of Hayami to investigate formation of acetol by decom-
position of hexoses via a hydrolytic cleavage of 1-deoxyhexo-2,4-
diulose (21). Also Weenen confirmed a β-cleavage mechanism to
give acetol and glyceric acid (22). Davidek postulated that first an
isomerization reaction of 1-DG to 1-deoxyhexo-2,4-diulose takes
place, followed-in the case of acetic acid-by a nucleophilic
attack of a hydroxyl ion at the C-2 carbonyl function (likewise
glyceric acid is formed by a nucleophilic attack on C-4 position).
The cleavage reaction results in acetic acid and a C4-enediole
which can undergo further isomerization reactions to yield
erythrulose and corresponding deoxyosones (7, 8, 20). These
findings give reason to believe that the cleavage mechanism will
Table 3. Comparison of Amide Concentrations Generated from 1-DG and
NR-t-BOC-lysine under Aerated and Deaerated Conditions
concentration (mmol/mol 1-DG)
24 h
72 h
aerated
deaerated
aerated
deaerated
Nε-acetyl lysine
Nε-formyl lysine
Nε-lactoyl lysine
Nε-glycerinyl lysine
0.48
1.12
0.19
0.11
0.45
0.71
0.16
0.05
0.78
1.47
0.25
0.15
0.75
1.08
0.26
0.09
Figure 3. Mechanism of nonoxidative 1-deoxyhexo-2,3-diulose degradation: β-dicarbonyl cleavage leads to Nε-acetyl lysine, Nε-glycerinyl lysine, Nε-lactoyl
lysine and Nε-formyl lysine.

Article
J. Agric. Food Chem., Vol. 58, No. 10, 2010 6463
structures 15 and 16 are formed after oxidation, respectively. This
step is followed by nucleophilic attack of a hydroxyl ion at the
tricarbonyls to give glyceric acid, lactid acid and formic acid as the
cleavage products. In our present study we measured significantly
higher amounts of Nε-formyl lysine and Nε-glycerinyl lysine under
aerated conditions. This is in line with the postulated mechanism,
because a nucleophilic attack of the ε-amino function of lysine
will lead to Nε-formyl lysine and Nε-glycerinyl lysine as shown
in Figure 4.
For verification that amides indeed originated from the pos-
tulated β-dicarbonyl cleavage reaction and not by a simple amide
bond formation between lysine side chains and the formed free
carboxylic acids, all these acids were separately incubated with
NR-t-BOC-lysine. Concentrations of carboxylic acids were cho-
sen according to the acetic acid concentration formed after seven
days in 1-DG/NR-t-BOC-lysine reaction mixtures. In all car-
boxylic acid/NR-t-BOC-lysine reaction mixtures the correspond-
ing amides were not detected. This control experimentthusclearly
depicts that the direct reaction of lysine with carboxylic acids is
definitively not an alternative pathway to produce the amides
under the given conditions.
€
While studies by Buttner and Hayase only provide an indica-
tion of a possible reaction between lysine side chains and sugars/
sugar degradation products (9, 10), we are now capable of
verifying Maillard induced amine acylation with the experiments
conducted herein. The detection of Nε-acetyl lysine, Nε-formyl
lysine, Nε-lactoyl lysine and Nε-glycerinyl lysine in Maillard
model reaction systems, resulting from reaction of the ε-lysine
side chain and 1-DG, opens a new field with respect to their
formation and significance in foods and in vivo. Compounds like
CML (Nε-(carboxymethyl)lysine) (25), CEL (Nε-(carboxyethyl)-
lysine) (26), GOLA (Nε-{2-[(5-amino-5-carboxypentyl)amino]-2-
oxoethyl}lysine), GALA (Nε-glycoloyllysine) and GOLD (1,3-bis-
(5-amino-5-carboxypentyl) imidazolium salt) (12) that occur via
rearrangement reactions have already confirmed that R-dicarbonyl
structures are highly potent to modify lysine side chains. Based on
our data it must be assumed that acylation induced by R-dicarbonyl
compounds is a second major pathway leading to a novel class of
amide advanced glycation end products (amide-AGEs) present in
food and in vivo.
Figure 4. Nucleophilic attack on tricarbonyls formed by oxidation from
1-DG or 1-DT leads to Nε-glycerinyl lysine and Nε-formyl lysine.
also lead to formation of carboxylic acid amides in the presence of
amines. In Figure 3 a mechanism is postulated that is capable of
elucidating the formation pathway of Nε-acetyl lysine, Nε-formyl
lysine, Nε-lactoyl lysine and Nε-glycerinyl lysine under deaerated
conditions. After isomerization of 1-DG (9) to 1-deoxyhexo-2,
4-diulose (10) a nucleophilic attack by the ε-amino function of
lysine at C-2 position of 10 followed by the β-cleavage reaction
results in the formation of Nε-acetyl lysine and C4-enediole (13).
In a similar manner the formation of Nε-glycerinyl lysine can be
explained by a nucleophilic attack of lysine at the C-4 position of
10, analogous to the formation of glyceric acid. Formation of
Nε-lactoyl lysine as well as formation of Nε-formyl lysine can
equally be explained by a β-cleavage mechanism assuming that
isomerization of 1-DG will also lead to 1-deoxyhexo-3,5-diulose
(11) and 1-deoxyhexo-4,6-diulose (12) (Figure 3). Such long-
range shifts of the carbonyl moiety in Maillard chemistry have
been already studied, leading to major intermediates, i.e. Lederer’s
glucosone (23, 24). Thus, isomerization of R-diketones into
β-diketones along the entire carbon backbone must be deemed
to be a general transformation.
In conclusion, we succeeded in the synthesis of Nε-lactoyl lysine
and Nε-glycerinyl lysine-lysine amide modifications that occur
in 1-DG/NR-t-BOC-lysine reaction mixtures. Furthermore we
established Nε-acetyl lysine, Nε-formyl lysine, Nε-lactoyl lysine
and Nε-glycerinyl lysine as novel stable end products in Maillard
reactions. The data reported in the present study strongly sup-
ports a formation mechanism of the amides based on the esta-
blished β-dicarbonyl cleavage reaction.
Strikingly, in comparison to the remainder carboxylic acid
amides, Nε-formyl lysine occurred with the highest amounts,
whereas formic acid gave the lowest values for all carboxylic
acids. With respect to formic acid formation this obviously
indicates that the nucleophilic attack of the ε-amino function of
lysine emerges as a competitive reaction to the hydrophilic attack
of the hydroxyl ion more strongly than in case of the other
carboxylic acid/amide pairs. This is based on the assumption that
β-dicarbonyl cleavage is the major formation pathway and an
additional oxidative R-dicarbonyl cleavage plays only a minor
part or is nonexistent (20).
Our working group has found a favored formation of glyceric
acid and lactic acid under aerated conditions and postulated a
formation mechanism coexistent to regular hydrolytic β-cleavage
to explain carboxylic acid formation (data will be published
elsewhere). Starting with 1-DG or 1-deoxy-2,3-tetrodiulose
(1-DT, a major degradation product of 1-DG (20)) the tricarbonyl
ACKNOWLEDGMENT
€
We thank Dr. D. Strohl from the Institute of Organic Chem-
istry, Halle, Germany, for recording NMR spectra and Dr. J.
Schmidt from the Leibnitz Institute of Plant Biochemistry, Halle,
Germany, for performing accurate mass determination.
LITERATURE CITED
(1) Hodge, J. E. Browning reactions in model systems. J. Agric. Food
Chem. 1953, 1, 928–943.
(2) Ledl, F.; Schleicher, E. The Maillard reaction in food and in the
human body - new results in chemistry, biochemistry and medicine.
Angew. Chem. 1990, 102, 597–626.
(3) Hirsch, J.; Mossine, V. V.; Feather, M. S. The detection of some
dicarbonyl intermediates arising from the degradation of Amadori
compounds (the Maillard reaction). Carbohydr. Res. 1995, 273,
171–177.

6464 J. Agric. Food Chem., Vol. 58, No. 10, 2010
Smuda et al.
(4) Hofmann, T. Quantitative studies on the role of browning precursors
in the Maillard reaction of pentoses and hexoses with L-alanine. Z.
Lebensm.-Unters. Forsch. A 1999, 209, 113–121.
Preparation of a Valinomycin Analog. J. Org. Chem. 1999, 64,
8063–8075.
(16) Glomb, M. A.; Tschirnich, R. Detection of alpha-dicarbonyl com-
pounds in Maillard reaction systems and in vivo. J. Agric. Food
Chem. 2001, 49, 5543–5550.
(17) Husek, P. Amino acid derivatization and analysis in five minutes.
FEBS Lett. 1991, 280, 354–356.
(18) Milne, G. W. A.; Axenrod, T.; Fales, H. M. Chemical ionization
mass spectrometry of complex molecules. IV. Amino acids. J. Am.
Chem. Soc. 1970, 92, 5170–5175.
(5) Oliver, C. M.; Melton, L. D.; Stanley, R. A. Creating proteins with
novel functionality via the Maillard reaction: A review. Crit. Rev.
Food Sci. Nutr. 2006, 46, 337–350.
(6) Glomb, M. A.; Pfahler, C. Synthesis of 1-deoxy-d-erythro-hexo-2,3-
diulose, a major hexose Maillard intermediate. Carbohydr. Res.
2000, 329, 515–523.
(7) Davidek, T.; Devaud, S.; Robert, F.; Blank, I. Sugar Fragmentation
in the Maillard Reaction Cascade: Isotope Labeling Studies on the
Formation of Acetic Acid by a Hydrolytic beta -Dicarbonyl Clea-
vage Mechanism. J. Agric. Food Chem. 2006, 54, 6667–6676.
(8) Davidek, T.; Robert, F.; Devaud, S.; Vera, F. A.; Blank, I. Sugar
Fragmentation in the Maillard Reaction Cascade: Formation of
Short-Chain Carboxylic Acids by a New Oxidative alpha -Dicarbonyl
Cleavage Pathway. J. Agric. Food Chem. 2006, 54, 6677–6684.
(9) Hayase, F.; Kim, S. B.; Kato, H. Maillard reaction products formed
from D-glucose and glycine and the formation mechanisms of amides as
major components. Agric. Biol. Chem. 1985, 49, 2337–2341.
(10) Buttner, U.; Gerum, F.; Severin, T. Formation of alpha -amino-acid
amides and alpha -hydroxy-acid amides by degradation of sugars
with primary amines. Carbohydr. Res. 1997, 300, 265–269.
(11) Nagaraj, R. H.; Shamsi, F. A.; Huber, B.; Pischetsrieder, M.
Immunochemical detection of oxalate monoalkylamide, an ascorbate-
derived Maillard reaction product in the human lens. FEBS Lett.
1999, 453, 327–330.
(19) Weinkam, R. J. Reactions of protonated diamino acids in the gas
phase. J. Org. Chem. 1978, 43, 2581–2586.
(20) Voigt, M.; Glomb, M. A. Reactivity of 1-Deoxy-D-erythro-hexo-2,3-
diulose: A Key Intermediate in the Maillard Chemistry of Hexoses.
J. Agric. Food Chem. 2009, 57, 4765–4770.
(21) Hayami, J. Studies on the chemical decomposition of simple sugars.
XII. Mechanism of acetol formation. Bull. Chem. Soc. Jpn. 1961, 34,
927–932.
(22) Weenen, H. Reactive intermediates and carbohydrate fragmentation
in Maillard chemistry. Food Chem. 1998, 62, 393–401.
(23) Biemel, K. M.; Conrad, J.; Lederer, M. O. Unexpected carbonyl
mobility in aminoketoses: the key to major Maillard crosslinks.
Angew. Chem., Int. Ed. 2002, 41, 801–804.
(24) Reihl, O.; Rothenbacher, T. M.; Lederer, M. O.; Schwack, W.
Carbohydrate carbonyl mobility;the key process in the formation
of alpha-dicarbonyl intermediates. Carbohydr. Res. 2004, 339, 1609–
1618.
(25) Reddy, S.; Bichler, J.; Wells-Knecht, K. J.; Thorpe, S. R.; Baynes,
J. W. Nɛ-(Carboxymethyl)lysine Is a Dominant Advanced Glycation
End Product (AGE) Antigen in Tissue Proteins. Biochemistry 1995,
34, 10872–10878.
(12) Glomb, M. A.; Pfahler, C. Amides are novel protein modifications
formed by physiological sugars. J. Biol. Chem. 2001, 276, 41638–
41647.
€
(13) Hasenkopf, K.; Ronner, B.; Hiller, H.; Pischetsrieder, M. Analysis of
(26) Ahmed, M.; Brinkmann Frye, E.; Degenhardt, T. P.; Thorpe, S. R.;
Baynes, J. W. Nε-(Carboxyethyl)lysine, a product of the chemical
modification of proteins by methylglyoxal, increases with age in
human lens proteins. Biochem. J. 1997, 324, 565–570.
glycated and ascorbylated proteins by gas chromathography-mass
spectrometry. J. Agric. Food Chem. 2002, 50, 5697–5703.
(14) Gellett, A. M.; Huber, P. W.; Higgins, P. J. Synthesis of the
unnatural amino acid NR-Nε-(ferrocene-1-acetyl)-
L-lysine: A novel
organometallic nuclease. J. Organomet. Chem. 2008, 693, 2959–2962.
(15) Kuisle, O.; Quinoa, E.; Riguera, R. A. General Methodology for
Automated Solid-Phase Synthesis of Depsides and Depsipeptides.
Received for review January 26, 2010. Revised manuscript received
April 14, 2010. Accepted April 17, 2010.