Assessment of the Antioxidative and Prooxidative Activities of Two Aminoreductones Formed during the Maillard Reaction: Effects on the Oxidation of β-Carotene, Nα-Acetylhistidine, and cis-Alkenes

Article abstract of DOI:10.1021/jf980118n

In short-time-heated mixtures of lactose and N^α-acetyllysine 1-[N^∈-(N ^α-acetyllysinyl)]-1,2-dehydro1,4-dideoxy-3-hexulose (C₆-AR) is formed as main product, whereas 3-hydroxy-4-(alkylamino)-3-buten-2-one (C

Full text of DOI:10.1021/jf980118n

J. Agric. Food Chem. 1998, 46, 2945−2950
2945
Assessm en t of th e An tioxid a tive a n d P r ooxid a tive Activities of Tw o
Am in or ed u cton es F or m ed d u r in g th e Ma illa r d Rea ction : Effects on
r
th e Oxid a tion of â-Ca r oten e, N -Acetylh istid in e, a n d cis-Alk en es
Monika Pischetsrieder,* Francesco Rinaldi, Ursula Gross, and Theodor Severin*
Institut f u¨ r Pharmazie und Lebensmittelchemie der Universit a¨ t M u¨ nchen, Sophienstrasse 10,
8
0333 M u¨ nchen, Germany
R
ꢀ
R
In short-time-heated mixtures of lactose and N -acetyllysine 1-[N -(N -acetyllysinyl)]-1,2-dehydro-
1
,4-dideoxy-3-hexulose (C6-AR) is formed as main product, whereas 3-hydroxy-4-(alkylamino)-3-
buten-2-one (C4-AR) can be obtained in high yields from the Maillard reaction of glucose. Because
both compounds have aminoreductone structure, their antioxidative (AOA) and prooxidative activities
(
POA) were determined and compared to those of ascorbic acid (AA). Concentration-dependent AOA
was determined by measuring oxidative degradation of carotene induced by a radical starter. POA
in the presence of metal ions was tested in three different systems: oxidation of carotene in emulsion,
of N -acetylhistidine in aqueous solution, and of cis-alkenes in organic solvent. C4-AR possesses in
R
all model systems AOA and POA, respectively, which are very similar to those of AA. C6-AR acts
also as antioxidant and prooxidant, but its activity is weaker compared to those of C4-AR and AA.
In the carotene assay the substances displayed POA in the presence of several metal ions, such as
2
+
2+
3+
Cu , Mn , and Fe /EDTA, but the activity with the latter is considerably lower.
Keyw or d s: Aminoreductones; Maillard reaction; antioxiative; prooxidative
INTRODUCTION
al., 1995), enzymes (Shinar et al., 1983), polysaccharides
Wong et al., 1981), DNA (Samuni et al., 1983), and
(
During the Maillard reaction reducing sugars react
with amino groups of amino acids or proteins and a
variety of products are formed depending on the reaction
conditions. In addition to other effects, Maillard prod-
ucts have strong reducing properties that can prevent
oxidative spoilage of processed foodstuffs, such as beer
lipids (Kanner et al., 1977) are oxidatively degraded.
Therefore, it has to be considered that Maillard products
in food and in vivo can also promote the oxidative
damage of other components. It must be assumed that
Maillard products can form H2O2 similarly to AA in the
presence of metal ions and oxygen, which then generates
in a Fenton reaction reactive oxygen species (ROS), such
as hydroxyl radicals.
(
Yoshimura et al., 1997; Nicoli et al., 1997; Karasto-
giannidou and Ryley, 1994.). Several Maillard products
with reductone structure that possess reducing proper-
ties have been identified (Ledl and Schleicher, 1990).
However, it is not clear how much of the total antioxi-
dative activity of Maillard products in food can be
assigned to these reductones. Because several studies
suggest that protein-bound Maillard products are re-
sponsible for scavenging reactive oxygen species (Alaiz
et al., 1997; Kato, 1992), other compounds apart from
reductones must be active.
When we investigated Maillard reaction mixtures of
disaccharides, 1-(alkylamino)-1,2-dehydro-1,4-dideoxy-
3
-hexulose (C6-AR) was obtained as the main product,
whereas 3-hydroxy-4-(alkylamino)-3-buten-2-one (C4-
AR) is formed in high yields when monosaccharides are
reacted (Scheme 1). Both the C6-AR from lactose and
the C4-AR from glucose possess a â-aminoreductone
structure (Scheme 1). It is known that products with a
reductone structure, such as AA, have reducing proper-
ties (Euler and Eistert, 1957). Furthermore, it was
deduced that the N-analogous R- and â-aminoreductones
have similar properties and that the electron-donating
effect of the amino group even increases the reducing
character (Pischetsrieder and Severin, 1997). However,
this assumption has not clearly been proven so far.
Therefore, the purpose of this study was to investigate
the prooxidant and antioxidant activities of C4-AR and
C6-AR in detail.
On the other hand, it is known that the majority of
antioxidants which are naturally present or added to
foodstuffs can also enhance free radical damage of other
components and therefore act as prooxidants in biologi-
cal systems (Aruoma, 1996). This process has been
thoroughly investigated for ascorbic acid (AA), and it
was found that in addition to its well-known antioxi-
dative properties, AA generates in the presence of metal
ions and oxygen reactive oxygen species, such as H2O2
•
(
Kalus et al., 1982), hydroxyl radicals ( OH) (Wong et
-
al., 1981), superoxide (O2 ) (Nakanishi et al., 1985), or
metal-peroxo complexes [e.g., O2-Cu(I)] (Uchida and
Kawakishi, 1989). As a result, peptides (Steinhart et
MATERIALS AND METHODS
Rea gen ts. AA was purchased from Roth (Karlsruhe,
Germany), and â-carotene, diethylenetriaminepentaacetic acid
(DTPA), and R,R′-azodiisobutyramidine dihydrochloride (ADI-
*
Author to whom correspondence should be addressed
telephone ++49-89-5902-387; fax ++49-89-5902-447; e-mail
pischets@pharmchem.uni-muenchen.de).
BA) were from Fluka (Buchs, Switzerland). Sodium linoleate
and Tween 20 were purchased from Sigma (St. Louis, MO).
Methanol LiChrosolv, chromatography grade, was obtained
(
S0021-8561(98)00118-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 07/07/1998

2946 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Pischetsrieder et al.
Sch em e 1. F or m a tion of C -AR a n d C -AR fr om D-Glu cose or La ctose, Resp ectively
4
6
from Merck (Darmstadt, Germany). Deionized water was
distilled before use for HPLC. Preparative thin-layer chro-
matography (TLC) was performed using 20 × 20 cm glass
methanol and separated by preparative HPLC on a LiChrosorb
RP 18, 250/2-7 µm column (Merck) with an eluent of 20%
methanol in 50 mM ammonium formate and a flow rate of 10
mL/min. The fractions between 30 and 40 min were collected,
unified, and extracted four times with 40 mL of ethyl acetate.
plates coated with a 0.5 mm thickness of silica gel 60 F254
.
Ap p a r a tu s. UV spectra were recorded on a Perkin-Elmer
1
UV-vis spectrometer Lambda 20 and H NMR (400 MHz) and
4
The unified organic layers were evaporated, and C -AR was
1
3
C NMR (100 MHz) spectra with a J EOL 400 GSX spectro-
purified by distillation at 150 °C and 0.4 Torr, which yields
1
meter using (CH
are reported in parts per million. Mass spectral analyses were
obtained with an HP 5989 A MS engine (CI with CH ) and
3
)
4
Si as internal standard. Chemical shifts
colorless crystals with a melting point of 47 °C: H NMR
(CDCl
(sext, 2 H, J ) 7.3 Hz, CH
3.11 (dt, 2 H, J ) 7.3 and 6.8 Hz, N-CH
3
, COSY) δ 0.90 (t, 3 H, J ) 7.3 Hz, CH
3
-CH
2
), 1.54
3
-CH ), 2.07 (s, 3 H, CH
2
3
CdO),
4
positive FABMS data with a Kratos MS 80 RFA spectrometer.
HP LC. Analytical HPLC was performed with a Merck
L-7100 gradient pump, a Merck L-7450 photodiode array
detector including Merck-Hitachi model D-7000 Chromato-
graphy Data Station software (Merck, Darmstadt). For quan-
tification DAD-System-Manager software D-7000 Chromato-
graphy Data Station (Merck-Hitachi) with manual baseline
correction was used. For preparative HPLC a Merck L-6250
pump, a Merck L-4000 UV detector, and a Merck D-2500
chromatointegrator were used. The mixtures were separated
on a column packed with Nucleosil 100-5 (RP 18, 125 × 3 mm
i.d., guard cartridge, 8 × 3 mm) from Macherey & Nagel
2
), 4.5 (br, 1 H, NH),
.57 (d, 1 H, J ) 12.4 Hz, CHdC); C NMR (CDCl ) δ 11.0
CH -CH ), 21.0 (CH -CdO), 24.5 (CH -CH ), 49.9 (CH
N), 131.0 (dC-OH), 131.7 (dCH-N), 186.4 (CdO); UV
1
3
6
3
(
3
2
3
3
2
2
-
+
(
CH
Anal. Calcd for C
C, 59.04; H, 8.93; N, 9.61.
Alternatively, 3-hydroxy-4-(propylamino)-3-buten-2-one (C
3
OH) λmax 321 nm (log ꢀ ) 4.3); CI-MS m/z 144 (M + 1).
7
H13NO
2
: C, 58.70; H, 9.16; N, 9.79. Found:
4
-
AR) was synthesized from 1-bromo-2,3-butanedione. Thus,
1-bromo-2,3-butanedione (165 mg, 1 mmol), which was pre-
pared according to the literature (Dow Chemical Co., 1958),
was dissolved in 1 mL of tetrahydrofuran and kept on ice. Two
millimoles of ice-cold propylamine was added dropwise to the
mixture and stirred for 30 min at room temperature. After
filtration, the solvent was evaporated and the residue was
separated by column chromatography on silica gel (solvent
(D u¨ ren, Germany). For elution a gradient was used of 0-100%
B from 0 to 25 min and 100% B from 25.1 to 40 min at a flow
rate of 0.5 mL/min (solvent A, 50 mM triethylammonium
acetate, pH 5.8; solvent B, methanol).
4
isopropyl ether/ethyl acetate 2:3, detection of C -AR with 2,6-
dichlorindophenone). Spectral data were identical with those
of the isolated product.
GC/MS. The GC/MS system consists of a Hewlett-Packard
(Waldbronn, Germany) gas chromatograph HP 5890 series II
coupled with a Hewlett-Packard HP 5971A msd mass spec-
trometer. Separations were performed on an Optima fused
silica capillary column 1701-0.25 µm (0.25 mm × 25 m;
Macherey & Nagel) with a helium flow of 0.5 mL/min (split:
An tioxid a n t Assa y. Antioxidant activity was assayed
according to a modified method of Chuda et al. (1996).
Solutions of â-carotene (0.5 mg in 0.5 mL), sodium linoleate
(20 mg in 0.2 mL), and Tween 20 (200 mg in 1 mL) in
chloroform were vigorously mixed, and the solvent was re-
moved under a stream of nitrogen. The mixture was dissolved
in 100 mL of water, and to 45 mL of this solution was added
4 mL of buffer (0.2 M phosphate buffer, pH 6.8, containing 5
mg/mL DTPA). The reagent solution (3.1 mL) was mixed with
1
/10). Oven temperature was programmed as follows for
detection of 2-cyclohexen-1-one: 4 min at 60 °C, from 60 to
8
2
2
4
2
0 °C at a rate of 2 °C/min, from 80 to 260 °C at a rate of
0 °C/min, and at 260 °C for 15 min. For the detection of
-hydroxy-3-hexene the program was as follows: 4 min at
0 °C, from 40 to 70 °C at a rate of 1 °C/min, from 70 to
60 °C at a rate of 20 °C/min, and at 260 °C for 15 min.
6 4
100 µL of the sample (various concentrations of C -AR and C -
Electron impact mass spectra were recorded under the follow-
ing conditions: capillary direct interface, 280 °C; ionization
voltage, 70 eV; mass range, m/z 50-400; electron multiplier
voltage, 2450 V; scan rate, 1.5 scans/s. Spectra were obtained
using HP G1034C MS ChemStation software.
AR in methanol and of AA in water as indicated), and the
reaction was started by addition of 5 µL of 0.5 M ADIBA in
phosphate buffer/DTPA. The blanks were prepared exactly
in the same way, but water or methanol was added instead of
sample. After 50 min of reaction in the dark, the absorbance
was measured at 470 nm. The antioxidative activity (AOA)
was calculated after subtraction of the blank as follows:
P r ep a r a tion of th e Am in or ed u cton es. 1-(Butylamino)-
1
6
,2-dehydro-1,4-dideoxy-3-hexulose (C -AR) was prepared as
described before (Pischetsrieder et al., 1997). Briefly, lactose
was heated with butylamine in phosphate buffer at pH 7.0
(Abs0min - Abs50min)/Abs0min.
The results are the mean of two to four independent
6
for 30 min at 100 °C, and C -AR was extracted with ethyl
experiments.
acetate. The analytically pure compound was immediately
used for the assays.
P r ooxidan t Assay Usin g â-Car oten e/Sodiu m Lin oleate.
Reagent solution was prepared as described above with the
exception that 0.2 M triethylammonium acetate, pH 7.0, was
used as buffer, unless otherwise noted. This reagent solution
(3.1 mL) was mixed with 100 µL of the sample (various
3
4
-Hydroxy-4-(propylamino)-3-buten-2-one (C -AR) was iso-
lated from a Maillard reaction mixture. Thirty milliliters of
an aqueous solution of propylamine (0.25 M) and glucose (0.25
M) was adjusted with phosphoric acid to pH 7.1 and was
heated for 1 h at 100 °C. The mixture was extracted three
times with 30 mL of ethyl acetate, and the solvent of the
organic layer was evaporated. The residue was dissolved in
concentrations of C
water as indicated), and the reaction was started by addition
of 8 µL of CuSO (3.2 mg/mL in water). As blanks, 100 µL of
methanol or water was used. Absorbance at 470 nm was
6 4
-AR and C -AR in methanol and of AA in
4

Antioxidative and Prooxidative Activities of Aminoreductones
J. Agric. Food Chem., Vol. 46, No. 8, 1998 2947
measured before and after 20 min of reaction in the dark.
Prooxidative activity (POA) was calculated as [Abs0min
-
Abs20min(sample)]/[Abs0min - Abs20min(blank)] × 100. The re-
sults are the mean of two independent experiments.
To determine the influence of different metal ions, instead
of CuSO
of FeCl
P r ooxid a n t Assa y Usin g N -Acetylh istid in e. CuSO
4
, 8 µL of MnCl
2 3
(4.0 mg/mL) or FeCl /EDTA (3.3 mg
3
and 7.6 mg of EDTA/mL) was used.
r
4
(16
µg/5 µL) was added to an aqueous solution of N -acetylhistidine
0.22 mg/mL), and the mixture was stirred vigorously in an
R
(
open vial. The reaction was started by the addition of the
sample (2 µmol/12 µL) and stirring was continued. After 30
min, phenoxyacetic acid (156 µg/200 µL) was added as an
internal standard and the solution was immediately injected
into the HPLC. Separation was performed on a column packed
with Lichrosorb 5C (RP 18, 250 × 4.6 mm i.d.) from Macherey
F igu r e 1. HPL-chromatogram of a reaction mixture of glucose
(0.25 M) and propylamine (0.25 M), which was heated in
phosphate buffer, pH 7.0, for 1 h at 100 °C. The substances
were detected in a wavelength range from 230 to 450 nm.
ether-heptane (3:2) (R
the silica gel with ethyl acetate, p-nitrobenzoic acid (3-hexen-
-yl) ester was obtained as colorless crystals: 1H NMR (CDCl
δ 1.01 (t, 3H, J ) 7.2 Hz, CH -CH ), 1.46 (d, 3H, J ) 6.1 Hz,
CH -CHO), 2.09 (quin, 2H, J ) 7.2 Hz, CH -CH ), 5.57 (m,
1H, CdCH-CHO), 5.61 (m, 1H, OCH-CH ), 5.87 (dt, 1H, J
5.9 and 14.6 Hz, CH -CHdC), 8.21 (d, 2H, 8 Hz, CHdC-
COO), 8.30 (d, 2H, 8 Hz, CHdC-NO ). The NMR data were
f
value ) 0.80), and after elution from
&
Nagel with an elution gradient of 0-25% B from 0 to 7 min
and 55-100% B from 7.1 to 30 min at a flow rate of 0.8 mL/
min (solvent A, 10 mM ammonium formate, pH 7.5; solvent
B, methanol). The substances were detected at 215 nm.
2
3
)
3
2
r
3
3
2
Id en tifica tion of N-F or m yl-N′-(N -a cetyl-â-a sp a r tyl)-
R
3
u r ea . N -Acetylhistidine (550 mg) and 725 mg of AA were
)
2
dissolved in 200 mL of water. After the addition of 9.6 mg of
CuSO in 3 mL of water, the mixture was stirred for 24 h at
4
2
in accordance with those which are reported in the literature
for acetic acid (3-hexen-2-yl) ester (Hansson et al., 1990).
room temperature and the excess of copper ions was removed
by an Amberlite IR 120 column. The eluate was lyophilized
and redissolved in 30 mL of water, filtered through a mem-
brane (0.2 µm pore size), and purified by preparative HPLC.
Separation was performed on a Supelcosil column LC-18-DB,
2
-Hydroxy-3-hexene: retention time, 6.9 min; MS, m/z (rela-
tive intensity) 85 (15%), 71 (100%), 69 (20%), 57 (20%), 55
(20%). 2-Cyclohexen-1-one: retention time, 6.9 min; MS, m/z
(relative intensity) 96 (35%), 68 (100%), 55 (10%), 51 (10%).
2
50 × 21.2 mm, 5 µm particle size (Supelco, Bellefonte, PA)
using water as eluent with a flow rate of 12 mL/min. Each
time 2 mL of the solution was injected and the compounds were
detected at a wavelength of 207 nm. The fractions that eluted
RESULTS AND DISCUSSION
between 11.3 and 12.9 min were collected, lyophilized, and
Previous studies have revealed that 1-(alkylamino)-
,2-dehydro-1,4-dideoxy-3-hexulose (C6-AR) is the main
1
subjected to spectral analyses: H NMR (DMSO-d
6
) δ 1.85 (s,
1
3
H, CH
COCH
CH
3
CO), 2.76-2.82 (dd, 1H, J ) 6.9 and 16.6 Hz,
), 2.94-3.00 (dd, J ) 5.9 and 16.6 Hz, 1H, CO-
), 4.58-4.64 (m, 1H, CH-CH ), 8.27-8.29 (d, 1H, J )
.0 Hz, NH-CH-CH ), 9.04 (d, 1H, J ) 9.5 Hz, NH-CHO),
0.51 (d, 1H, J ) 9.5 Hz, NH-CHO), 11.02 (s, 1H, CO-NH-
product, which can be detected by HPLC/UV, when
a
H
b
R
lactose is heated with alkylamines, such as N -acetyl-
a
H
b
2
8
1
2
lysine (Pischetsrieder et al., 1997). 3-Hydroxy-4-(alkyl-
amino)-3-buten-2-one (C -AR) has previously been iso-
4
1
CO); H NMR (DMSO-d
.82 (dd, 1H, J ) 6.9 and 16.6 Hz, COCH
J ) 5.9 and 16.6 Hz, 1H, COCH ), 4.58-4.61 (dd, 1H, J )
.1 and 6.8 Hz, CH-CH ), 9.04 (s, 1H, NH-CHO); FABMS
Xe, 7 kV, glyerol), m/z 246 [M + H]; UV (H O) λmax 206 nm.
P r ooxid a n t Assa y Usin g cis-Hexen es. To 3 mL of
acetonitrile (HPLC grade, Merck) were added 0.1 mmol of C
AR, AA, or 5,6-O-isopropylideneascorbic acid (Micheel and
Hasse, 1936), 150 µL of a solution of 10% FeCl in water, and
.6 mmol of cyclohexen or cis-3-hexene, respectively. Air was
6
+ D
2
O) δ 1.85 (s, 3 H, CH
3
CO), 2.76-
lated from reaction mixtures of glucose and alkylamines,
but the yields have been rather low (Ledl and Severin,
979). In course of these experiments HPLC analysis
2
a b
H
), 2.94-3.00 (dd,
a
H
b
1
6
2
was used to investigate the Maillard reaction of glucose
in detail, and it was found that under certain reaction
conditions, such as heating for 1 h at 100 °C and neutral
pH, C4-AR is the major UV-absorbing product which
could be detected (Figure 1).
To investigate the AOA of C4-AR and C6-AR, a linoleic
acid/carotene mixture was oxidized in the absence of
metal ions by the addition of ADIBA, which starts
radical reactions, and the activity of the samples to
prevent carotene degradation was measured. The re-
sults are summarized in Figure 2. It was found that
the AOA of C4-AR is over a wide concentration range
almost the same as that of AA, and an antioxidative
effect can even be observed in concentrations as low as
(
2
4
-
3
0
bubbled through the mixture for 100 min with stirring. After
the reaction, 50 µL of cycloheptanol in methanol (12.1 mg/mL)
was added as internal standard, and the organic solvent was
removed under reduced pressure at room temperature. The
residue was diluted with 3 mL of ethyl acetate and dried over
2 4
Na SO anhydrous, and the solution was filtered through a
membrane of 0.45 µm pore size (Chromafil, Macherey &
Nagel). The products were identified by comparing of reten-
tion time and mass spectrum of those of the authentic
reference compound. 2-Cyclohexen-1-one was obtained from
Fluka, and 2-hydroxy-3-hexene was synthesized according to
a modified method of Kothe et al. (1994). Selenium dioxide
0
.015 mM. C6-AR is a weaker antioxidant, which is still
effective in a concentration of 0.02 mM.
It is known that in the presence of metal ions
antioxidants can be effective as prooxidants against
hydrophilic compounds, such as proteins, and lipophilic
compounds, such as unsaturated lipids. Therefore, we
(3.5 mmol) was suspended in 4 mL of methylene chloride and
1
.6 mL of 80% tert-butyl hydroperoxide in bis(tert-butylper-
oxide), 0.08 mL of water, and 0.08 mL of tert-butyl alcohol were
added. cis-3-Hexene (3.5 mmol) was added to the stirred
reaction mixture. After 24 h of further stirring, the solvent
was evaporated, and the residue was suspended in 5 mL of
water and extracted three times with 5 mL of ethyl acetate.
After the pooled organic layers were evaporated, an oil was
obtained that contained 2-hydroxy-3-hexene as identified by
GC/MS. For further characterization of the product, the oil
was derivatized for 15 min at room temperature in an excess
of p-nitrobenzoyl chloride in pyridine. The product was
isolated by preparative TLC with an eluent of diisopropyl
R
developed two assay systems, one using N -acetylhis-
tidine as a model for proteins and one using linoleate/
R
carotene as an example for lipids. N -Acetylhistidine
was used because it is known that particularly histidine
residues are prone to degradation during the metal ion-
mediated oxidation of proteins and peptides (Cheng et
al., 1992). It was suggested that histidine forms a
complex with copper, which is directly reduced by the

2948 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Pischetsrieder et al.
F igu r e 2. AOA dependent on the concentration of AA (s),
-AR (9), and C -AR (2) tested for the ADIBA-induced
oxidation of carotene. Concn, concentration of the sample
which was added to the test solution.
C
4
6
F igu r e 4. POA dependent on the concentration of AA (s),
C -AR (9), and C -AR (2) tested for the Cu -induced oxidation
4 6
of carotene. Concn, concentration of the sample which was
added to the test solution.
2+
between 0.005 and 0.1 mM sample. In concentrations
>
0.1 mM the AOA of both compounds becomes pre-
dominant, resulting in a decrease of POA. The effect
that AA enhances at lower concentrations oxidation but
acts as an antioxidant at high concentrations has been
described in the literature (Steinhart et al., 1993;
Mahoney and Graf, 1986). The values for the C6-AR
were again significantly lower.
Because Fenton-type reactions were described for
various metal ions, we tested if C4-AR and C6-AR can
display their POA also in the presence of other metals.
Therefore, in the linoleate/carotene assay copper was
replaced by MnCl2 or FeCl3/EDTA. The samples dis-
played POA in the presence of all tested metal ions. For
R
F igu r e 3. Oxidative degradation of N -acetylhistidine in the
2
+
2+
2
+
Mn results similar to those for Cu were obtained,
4 6
presence of Cu and AA, C -AR, or C -AR.
3
+
whereas Fe caused less than half of the POA.
antioxidant and reacts with oxygen to give a strongly
oxidizing oxo-copper species (Uchida and Kawakishi,
It is known that AOA or POA against lipids can be
highly dependent on the model system chosen (Hopia
et al., 1996), particularly if the lipids were used in
emulsions, bulk systems, or organic solvents. Therefore,
a second assay was used to determine the POA of C4-
1
990). However, in another study it was suggested that
H2O2 plays a role in the oxidation process (Cheng et al.,
992). We developed an HPLC system to monitor the
1
R
3+
degradation of N -acetylhistidine during the reaction
AR against lipids in the presence of Fe . cis-Alkenes,
2
+
with C4-AR or C6-AR, Cu , and O2. To confirm that
degradation of N -acetylhistidine is due to oxidation and
such as cyclohexene and cis-3-hexene, were used as
model compounds for oxidation of oleic acid in organic
solvent. The reaction products were analyzed by GC/
MS and identified by comparing retention times and
mass spectra with those of authentic reference com-
pounds. It was found that oxidation in the R-position
of cyclohexene leads to the formation of 2-cyclohexen-
1-one, whereas the open chain alkene is oxidized to give
2-hydroxy-3-hexene (Scheme 2). The reactivity was
again compared to this one of AA which was used
directly in suspension or as 5,6-isopropylidene deriva-
tive, which is soluble in organic solvent. It was found
that AA, 5,6-isopropylidene AA, and C4-AR produce the
same reaction products in similar yields. Very recently,
Barton and Delanghe (1998) showed that cyclohexanone
is in a similar way the main product when cyclohexane
is oxidized by an Udenfriend’s system.
The mechanism of the POA of aminoreductones has
not been elucidated so far and seems to be dependent
on the reaction conditions. It can be suggested, how-
ever, that in the first step the aminoreductone forms a
reactive complex with the metal and oxygen (Scheme
2). The analogous complex with AA was described by
Hamilton as the oxidant of Udenfriend systems (Hamil-
ton, 1974; Elstner, 1990; Uchida and Kawakishi, 1990).
Because of the similar structure and oxidative behavior
R
not due to other processes, the main reaction product,
which could be detected by HPLC, was isolated and
R
identified as N-formyl-N′-(N -acetyl-â-aspartyl)urea. This
compound has been previously postulated as an inter-
mediate of the metal ion/AA-induced oxidative degrada-
R
tion of N -acetylhistidine, but it has not been isolated
so far (Uchida and Kawakishi, 1990). However, since
several products, including aspartate and asparagine,
are known to be formed during this reaction, degrada-
tion of the educt, and not product formation, has been
quantified in this assay.
The results were compared to AA, which is known to
be a very potent prooxidant in the presence of metal
ions. Copper alone did not show significant reaction,
and this sample was therefore used as blank.
As shown in Figure 3, it was found that C6-AR causes
R
the degradation of 4.6% of N -acetylhistidine. On the
other hand, C4-AR was much more effective, causing a
loss of 31.8% of the amino acid, which is comparable to
the result of AA (36.4%).
In the linoleate/carotene system results similar to
R
those for the N -acetylhistidine test were obtained
(
Figure 4). The dose-dependent curve for C4-AR is very
similar to the curve for AA, with increasing POA

Antioxidative and Prooxidative Activities of Aminoreductones
J. Agric. Food Chem., Vol. 46, No. 8, 1998 2949
r
Sch em e 2. Oxid a tion of cis-Alk en es a n d N -Acetylh istid in e in th e P r esen ce of Am in or ed u cton es a n d Meta l Ion s
of reductones, such as AA, and aminoreductones, it is
likely that the complex indicated in Scheme 2 can be
formed and can be an intermediate of the oxidation
reactions. This complex can then either oxidize directly
the substrate (Elstner, 1990), or superoxide radicals can
be released. After dismutation of the latter and a
Fenton reaction, H2O2 and OH can be formed as
reactive oxygen species (ROS). Against our expecta-
tions, oxidation of cis-alkenes could be observed only
damage of DNA (Kalus et al., 1982) or LDL (J u¨ rgens et
al., 1987). In these cases aminoreductones, which are
formed during the Maillard reaction, can be of particular
interest, because they are mostly bound to proteins and
the close proximity of the reducing agent can greatly
influence the oxidative damage of proteins and lipopro-
teins. As indicated above, AA influences oxidation of a
wide range of molecules, including DNA and phospho-
lipids. Since AA and the aminoreductones showed very
similar behaviors in the here-applied assay systems, it
can be deduced that C4-AR and C6-AR can have anti-
or prooxidative activity on other components such as
phospholipids and DNA. However, further experiments
are needed to confirm this assumption.
•
3
+
2+
when Fe was used, whereas Fe did not show an
effect. This observation can possibly be explained by
the well-known fact that the tendency of Fe3 to form
+
2
+
complexes is much higher compared to that of Fe . It
is possible that synergistic effects of AA and C4-AR occur
in this oxidation system; however, this was not inves-
tigated in the scope of this project.
Further investigations to find out to what extent C4-
AR and C6-AR are formed in foodstuffs or in vivo are
currently under progress.
In summary, it can be stated that C4-AR and C6-AR
are strongly reducing agents which have, dependent on
the conditions, both antioxidative and prooxidative
properties. In all assay systems the activity of C4-AR
was in the range of the activity of AA and significantly
higher than that of C6-AR. It must be assumed that
both aminoreductones contribute considerably to the
anti- and prooxidative effect of Maillard products. On
the one hand, these aminoreductones can help to
prevent oxidative spoilage of foodstuffs. On the other
hand, at low ratios of aminoreductone to metal ions, the
aminoreductones can enhance oxidative damage of
proteins, lipids, or other components. In vivo under
normal conditions metal ions are sequestered. However,
there is strong evidence that under pathological condi-
tions, such as cataract (Cook and McGahan, 1986),
atherosclerotic lesions (Evans et al., 1995; Smith et al.,
ABBREVIATIONS USED
AA, ascorbic acid; ADIBA, R,R’-azodiisobutyramidine
dihydrochloride; AOA, antioxidative activity; C4-AR,
3
-hydroxy-4-(alkylamino)-3-buten-2-one; C6-AR, 1-(al-
kylamino)-1,2-dehydro-1,4-dideoxy-3-hexulose; DAD, di-
ode array detection; DTPA, diethylenetriaminepentaace-
tic acid; EDTA, ethylenediaminetetraacetic acid; HPLC,
high-performance liquid chromatography; POA, prooxi-
dative activity.
LITERATURE CITED
Alaiz, M.; Hidalgo, J .; Zamora, R. Comparative Antioxidant
Activity of Maillard-and Oxidized Lipid-Damaged Bovine
Serum Albumin. J . Agric. Food Chem. 1997, 45, 3250-3254.
1
992), and diabetes (Hunt, 1996), metal ions are re-
leased and can be detected in their free and harmful
form. Very recently, it was shown that AA exhibits POA
also in vivo, resulting in an increase of potentially
mutagenic lesions (Podmore et al., 1998). In the pre-
sence of reducing agents metals can cause, for example,
inactivation of enzymes (Stadtman and Oliver, 1991) or
Aruoma, O. Assessment of Potential Prooxidant and Antioxi-
dant Actions. J . Am. Oil Chem. Soc. 1996, 73, 1617-1625.
Barton, D. H.; Delanghe, N. C. The Selective Functionalization
of Saturated Hydrocarbons. Part 46. An Investigation of
Udenfriend’s System under Gif Conditions. Tetrahedron
1998, 54, 4471-4476.

2950 J. Agric. Food Chem., Vol. 46, No. 8, 1998
Pischetsrieder et al.
Cheng, R.-Z.; Uchida, K.; Kawakishi, S. Selective Oxidation
of Histidine Residues in Proteins or Peptides through the
Copper(II)-Catalysed Autoxidation of Glucosone. Biochem.
J . 1992, 285, 667-671.
Chuda, Y.; Ono, H.; Mayumi, O.-K.; Nagata, T.; Tsushida, T.
Structural Identification of Two Antioxidant Quinic Acid
Derivatives from Garland (Chrysanthmum coronarium L.).
J . Agric. Food Chem. 1996, 44, 2037-2039.
Mahoney, J . R.; Graf, E. Role of R-tocopherol, ascorbic acid,
citric acid and EDTA as oxidants in model systems. J . Food
Sci. 1986, 51, 1293-1296.
Micheel, F.; Hasse, K. U¨ ber die 2-Deoxy-L-ascorbins a¨ ure.
Chem. Ber. 1936, 69, 879.
Nakanishi, Y.; Isohashi, F.; Matsunaga, T.; Sakamoto, Y.
Oxidative Inactivation of an Extramitochondrial Acetyl-CoA
Hydrolase by Autoxidation of L-Ascorbic Acid. Eur. J .
Biochem. 1985, 152, 337-342.
Cook, C. S.; McGahan, M. C. Copper concentration in cornea,
iris, normal, and cataractous lenses and intraocular fluids
of vertebrates. Curr. Eye Res. 1986, 5, 69-76.
Nicoli, M. C.; Anese, M.; Parpinel, M. T.; Franceschi, S.; Lerici,
C. R. Loss and/or formation of antioxidants during food
processing and storage. Cancer Lett. 1997, 114, 71-74.
Pischetsrieder, M.; Severin, T. New Aspects on the Maillard
ReactionsFormation of Aminoreductones from Sugars and
L-Ascorbic Acid. Recent Res. Dev. Agric. Food Chem. 1997,
1, 29-37.
Dow Chemical Co. U.S. Patent 282/555; Chem. Abstr. 1958,
5
2, 9198i.
Elstner, E. Der Sauerstoff. BI-Wissenschaftsverlag: Mann-
heim, 1990.
Euler, H.; Eistert, B. Chemie und Biochemie der Reduktone
und Reduktonate; Ferdinand Enke Verlag: Stuttgart, 1957.
Evans, P. J .; Smith, C.; Mitchinson, M. J .; Halliwell, B. Metal
ion release from mechanically-disrupted human arterial
wall. Implications for the development of artherosclerosis.
Free Radical Res. 1995, 23, 465-469.
Hamilton, G. A. In Molecular Mechanisms of Oxygen Activa-
tion; Hayaishi, O., Ed.; Academic Press: New York, 1974;
p 405.
Hansson, S.; Heumann, A.; Rein, T.; Akermark, B. Preparation
of Allylic Acetates from Simple Alkenes by Palladium (II)-
Catalyzed Acetoxylation. J . Org. Chem. 1990, 55, 975-984.
Hopia, A. I.; Huang, S.-W.; Schwarz, K.; German, J . B.;
Frankel, E. N. Effect of Different Lipid Systems on Anti-
oxidant Activity of Rosemary Constituents Carnosol and
Carnosic Acid with and without R-Tocopherol. J . Agric. Food
Chem. 1996, 44, 2030-2036.
Hunt, J . V. Ascorbic Acid and Diabetes Mellitus. In Subcellular
Biochemistry, Vol. 25: Ascorbic Acid: Biochemistry and
Biomedical Cell Biology; Harris, J . R., Ed.; Plenum Press:
New York, 1996.
Pischetsrieder, M.; Schoetter, C.; Severin, T. Formation of an
Aminoreductone during the Maillard Reaction of Lactose
R
with N -Acetyllysine or Proteins. J . Agric. Food Chem. 1998,
46, 928-931.
Podmore I. D.; Griffiths, H. R.; Herbert, K. E.; Mistry, N.;
Mistry, P.; Lunec, J . Vitamin C exhibits pro-oxidant proper-
ties. Nature 1998, 392, 559.
Samuni, A.; Aronovitch, J .; Godinger, D.; Chevion, M.; Czapski,
G. On the Cytotoxicity of Vitamin C and Metal Ions. A Site-
Specific Fenton Mechanism. Eur. J . Biochem. 1983, 137,
119-124.
Shinar, E.; Navok, T.; Chevion, M. The Analogous Mechanism
of Enzymatic Inactivation Induced by Ascorbate and Su-
peroxide in the Presence of Copper. J . Biol. Chem. 1983,
258, 14778-14783.
Smith, C.; Mitchinson, M. J .; Aruoma, O. I.; Halliwell, B.
Stimulation of lipid peroxidation and hydroxyl-radical gen-
eration by the contents of human atherosclerotic lesions.
Biochem. J . 1992, 286, 901-905.
Stadtman, E. R.; Oliver, C. N. Metal-catalyzed Oxidation of
Proteins. J . Biol. Chem. 1991, 266, 2005-2008.
Steinhart, H.; Vollmar, M.; Sailer, C. Pro-and Antioxidative
J u¨ rgens, G.; Hoff, H. F.; Chisolm, G. M:, Esterbauer, H.
Modification of Human Serum Low-Density Lipoprotein by
OxidationsCharacterization and Pathophysiological Impli-
cations. Chem. Phys. Lipids 1987, 45, 315-336.
Kalus, W.; Filby, W.; M u¨ nzner, R. Chemical Aspects of the
Mutagenic Activity of the Ascorbic Acid Autoxidation Sys-
tem. Z. Naturforsch. 1982, 37C, 40-45.
3
+
Effect of Ascorbic Acid on L-Tryptophan in the System Fe
/
Ascorbic Acid/O . J . Agric. Food Chem. 1993, 41, 2275-2277.
2
Steinhart, H.; Meyer, K.; Vollmar, M. Stability of Tryptophan-
Containing Peptides in the Presence of an L-Ascorbic Acid-
Ferric Ion System. J . Agric. Food Chem. 1995, 43, 2321-
2324.
Kanner, J .; Mendel, H.; Budowski, P. Prooxidant and Anti-
oxidant Effects of Ascorbic Acid and Metal Salts in a
â-Carotene-Linoleate Model System. J . Food Sci. 1977, 42,
Uchida, K.; Kawakishi, S. Ascorbate-Mediated Specific Modi-
fication of Histidine-Containing Peptides. J . Agric. Food
Chem. 1989, 37, 897-901.
6
0-64.
Karastogiannidou, C.; Ryley, J . The Formation of Water-
soluble Antioxidants in Chicken Held at 80 Degree C. Food
Chem. 1994, 51, 215-220.
Kato, H. Scavenging of Active Oxygen Species by Glycate
Protein. Biosci., Biotechnol., Biochem. 1992, 56, 928-931.
Kothe, G.; Bagby, M. O.; Weisleder, D.; Peterson, R. E. Allylic
Mono- and Di-hydroxylation of Isolated Double Bonds with
Selenium Dioxide-tert-Butyl Hydroperoxide. NMR Charac-
terization of Long-chain Enols, Allylic and Saturated 1,4-
Diols, and Enones. J . Chem. Soc., Perkin Trans. 2 1994,
Uchida, K.; Kawakishi, S. Formation of the 2-Imidazolone
Structure within a Peptide Mediated by a Copper(II)/
Ascorbate System. J . Agric. Food Chem. 1990, 38, 1896-
1899.
Wong, S. F.; Halliwell, B.; Richmond, R.; Skowroneck, W. R.
The Role of Superoxide and Hydoxyl Radicals in the
Degradation of Hyaluronic Acid Induced by Metal Ions and
by Ascorbic Acid. J . Inorg. Biochem. 1981, 14, 127-134.
Yoshimura, Y.; Iijima, T.; Watanabe, T.; Nakazawa, H. Anti-
oxidative Effects of Maillard Reaction Products Using
Glucose-Glycine Model System. J . Agric. Food Chem. 1997,
45, 4106-4109.
1
661-1669.
Ledl, F.; Schleicher, E. New aspects of the Maillard reaction
in foods and in the human body. Angew. Chem., Int. Ed.
Engl. 1990, 29, 565-594.
Received for review February 10, 1998. Revised manuscript
received May 19, 1998. Accepted May 19, 1998.
Ledl, F.; Severin, T. Formation of Aminoreductones from
Glucose and Primary Amines. Z. Lebensm. Unters. Forsch.
1
979, 169, 173-175.
J F980118N