close

Вход

Забыли?

вход по аккаунту

?

New Aspects of the Maillard Reaction in Foods and in the Human Body.

код для вставкиСкачать
Volume 29 . Number 6
J u n e 1990
Pages 565-706
International Edition in English
New Aspects of the Maillard Reaction in Foods
and in the Human Body
By Franz Ledl* and Erwin Schleicher"
Translated by Harry E. Nursten
Dedicated to Professor Otto H. Wieland on the occasion of his 70th birthday
and to Projessor Theodor Severin on the occasion of his 60th birthday
The reactions that occur during the cooking, baking, and preservation of foods of all kinds are
of great importance for the production of aroma, taste, and color. However, more recently it
has been shown that these reactions may be accompanied by a reduction in nutritive value and
the formation of toxic compounds. For these reasons, the very complex reactions between
reducing sugars and the free amino groups of amino acids or proteins, known as non-enzymatic browning or the Maillard reaction, have again caught the interest of chemists. The Maillard
reaction came to be seen in a new light as it was realized that it actually occurred in the human
body. As a general rule, the longer the half-life of a protein, the larger the amount of its
Maillard products found, i.e., important factors are the 'age' or persistence of the protein in
the body and the glucose concentration, particularly in diabetics. Many of the symptoms
developed by diabetics resemble those of premature aging, which leads to the possibility that
glucose, because of its reactivity towards proteins, is fundamentally involved in the normally
slow progress of aging.
1. Introduction
The first substantial investigation of the reaction of reducing sugars with amino acids was carried out about 75 years
ago by Maillard.['' Subsequently, Arnadori''] reported the
formation of two reaction products, which he described as
the stable and the unstable isomer. Kuhn and Weygand
['I
Prof. Dr. F. Led1
Institut fur Lebensmittelchemie und Anaiytische Chemie der Universitat
Pfaffenwaldring 55, D-7000 Stuttgart 80 (FRG)
Dr. rer. nat. habii. E. Schleicher
Institut fur Klinische Chemie und Institut fur Didbetesforschung am
Stadtischen Krankenhaus Miinchen-Schwabing
Kolner Piatz 1, D-8000 Munchen 40 (FRG)
Angen,. Chem. In(. Ed. En$. 29 (1990) 565-594
8 VCH
showed later that a glycosylamine (the labile isomer) is
formed first and that it then rearranges to the aminoketose
(the stable isomer).[3JThe elucidation of the mechanism of
the reaction intrigued Simon and Kraus.14' Heyns et al.l5]
reported on the corresponding reactions of ketoses with amino acids that result in the formation of aminoaldoses. Further contributions to an understanding of the 'early' stage of
the Maillard reaction, both with regard to the reactions occurring and to their products, came amongst others from the
schools of Hodge, Anet, Heyns, Micheel, and Kato.I6' In the
following sections, the significance of the Maillard reaction
for foods and for the human organism will be outlined on the
basis of the more recently achieved results.t71
Verlags&?e.sellschafr mhH, 0-6940 Weinheim, i990
oS70-0833~90f060~-0565
S 0 3 3 + .25/0
565
2. Reaction Paths
2.1. Preamble
The Maillard reaction, or non-enzymatic browning, embraces not only one reaction but a whole network of different
reactions. Accordingly, when sugars react with amines, one
obtains an extraordinarily complex mixture of compounds,
which are present in very different amounts. Besides substances which may be formed in relatively high yields (up to
30%) under certain conditions, other compounds are
formed whose concentrations lie in the ppb range or even
lower. Contrary to preparative chemistry, here the significance of a reaction and of a product is determined not so
much by the yield but by its effect (e.g., aroma, taste). A few
pg of a product may well be sufficient to characterize the
aroma of a heated food. The product pattern is subject to
large variations, depending on the reaction conditions. In
particular, the reaction time, the temperature, the concentrations of the reactants, and the pH play a decisive role. These
factors are responsible for the differences in aroma, taste,
and appearance of boiled meat and roasted meat.
Although more than 75 years have passed since the first
research on the Maillard reaction and although many results
have been gathered in the meantime, it is still not possible to
present a complete reaction scheme. To date, only a few of
the reaction products have been isolated and identified. The
products that have been characterized are preponderantly
those that are stable and have not undergone further change
to any large extent either in the reaction mixture or during
isolation or purification. Such compounds can serve as indicator substances for certain reaction paths in the Maillard
reaction. Difficulties are encountered in the isolation O f redCtive intermediates, since they are present only in very low
concentrations in the reaction mixture and they usually react
further during isolation. Yet it is precisely these compounds
that are of the greatest significance for the Maillard reaction,
since they play an important role in the formation of browning products, aroma compounds, high molecular weight substances, and so on.
In the following, an attempt will be made to describe the
reaction sequence of the Maillard reaction without going
into details that can often be unnecessarily confusing. Consequently the results of some studies are not or only partly
taken into account, without any intention of detracting from
their importance.
2.2. Initial Reactants
In foods, it is essentially the monosaccharides, glucose and
fructose, and the disaccharides, maltose and lactose, as well
as in some cases (e.g., meat) reducing pentoses, that react
with amino acids and/or proteins. Sugar linked glycosidically in glycoproteins, glycolipids, flavonoid compounds, or in
disaccharides, such as sucrose, participate in the Maillard
reaction only after cleavage of the glycosidic bond. Hexuronic acids, insofar as they occur in the free form at all, behave
like pentoses, i.e., in the course of the Maillard reaction
decarboxylation takes place. In certain cases (e.g., cheese),
biogenic amines can react as the amino component. Ammonia constitutes a special case; it can be formed in small
amounts from amino acids during the Maillard reaction (cf.
Section 2.7.1) and it i s used in large amounts in the production of one type of caramel coloring.
In the human body, the reaction between glucose and the
free amino groups of proteins as the amino component has
been the primary subject of investigation. More recent results have indicated that fructose and pentoses also play a
role as sugar components (cf. Sections 4.2 and 4.5). In general, for the Maillard reaction, primary amines are more im-
Franz Led1 was born in Starnberg, Germany in 1946. He studied food chemistry at the University
of Munich where he received his Dr. rer. nut in 1973 under the supervision of Th. Severin and
habilitated in 1987. At present he is Projessor of Food Chemistry and Analytical Chemistry at
Stuttgart University. His research interests include several areas of the Maillard reaction infoods
and in vivo.
Erwin Schleicher was born in Bamberg in 1946. He studied chemistry in Munich at the Technical
University from 1968- 1972. After graduation he remained in Munich studying under the supervision of H. Simon and receiving his doctorate in 1974. He spent the,following years (19751976) carrying outpostdoctoral research with H. G. Floss at Purdue University (Indiana, U S A ) .
After returning to Munich he investigated the mechanism of redox enzymes. Since 1977 he has
been working as a clinical chemist at the institutes for clinical chemistry and diabetes research
group at the academic hospital Munich-Schwabing with 0. H. Wieland. In 1986 he habilitated at
TU Munich. His interests revolve around the theme of structurallfunctional relationship of
glycoproteins, especially extracellular proteins, and the mechanism of signal transmission
through cell membranes.
566
Angen. Chem. Ini. Ed Engl. 29 (1990) 565-594
portant than secondary amines. Thus, in proteins, it is the
primary amincl groups of the lysine side-chains that react
predominantly and, insofar as they occur in foods in the free
form. primary amino acids take precedence. However, in
cereals and in cereal products (malt, beer), considerable
amounts of the secondary amino acid proline occur; since
cereals are staple items of the diet, attention will also be paid
to the behavior of secondary amines in the Maillard reaction.
2.3. Model Reactions
Investigations of the Maillard reaction in foods and also in
vivo have been carried out predominantly with the help of
model systems. The main reason for this is to be sought in the
wide range of reactants and other substances present in even
the simplest of natural systems. Not only do they make the
situation that much more complex, but they make the separation, isolation, and purification of Maillard products extremely difficult. Fortunately, it has become evident that the
relationships and conditions in foods and in the human organism can be simulated with model systems and that the
results obtained can be extrapolated. The knowledge presented in Sections 2.4 to 2.8 on the reaction pathways has
been largely obtained from such model systems.
organic solvent. The non-polar products that are extracted
into the organic phase can be analyzed gas chromatographically, provided they are sufficiently volatile, and characterized mass spectrometrically. In recent years, this has been the
method by which most Maillard products have been identified. After derivatization (silylation, acylation), some of the
polar products can also be identified gas chromatographically. The remaining mixture, however, presents great difficulties for the analyst, yet it constitutes the major part of the
reaction products. By conventional chromatographic methods (exclusion, ion exchange, adsorption, partition), fractions are first obtained, which may yield pure substances, but
usually only after several separation steps. High-performance liquid chromatography (HPLC) has provided reproducible fractionation and has facilitated the recognition of
substances and of class of substance through the use of different types of detector (UV, fluorescence, electrochemical,
etc.). To answer more detailed questions regarding a particular reaction mechanism or the elucidation of a new reaction
path, starting substances labeled with radioactive and/or stable isotopes have been used. In the following sections, the
analyses of some Maillard products are detailed and examples are given of derivatkation and trapping reagents. Some
problems in the investigation of protein-sugar reaction mixtures are outlined in Section 4.3.
2.3.1. Types of Model Systems
In general, in order to simulate conditions in foods and in
vivo, aqueous solutions of the sugar and the amino component have been investigated. To study the processes occurring during roasting, fusion of the reactants has been found
appropriate. Bearing in mind that the sugars and also the
amines occur in foods in widely different amounts, the number of possible model systems is clearly vast. Yet care must be
taken, since several preparative operations involve concentration and it is known that, for example, in drying foods, a
range of concentrations has to be traversed that is optimal
for the Maillard reaction to occur.
Ofcentral significance for product formation are time and
temperature (increases of up to sixfold have been recorded).
As a rule of thumb, the rate at least doubles when the temperature is increased by 10 "C. If browning is used to measure the progress of the Maillard reaction, then four weeks at
20 "C, three hours at 1OO"C, or five minutes at 150°C give
approximately the same result.
In the main, model systems are buffered to a pH between
4 and 7.5.
To simulate the conditions in the human or other animal
organism, the sugar component, glucose, is incubated with
the amine component, e.g., protein, 140 mM sodium chloride, and 5 mM potassium chloride at pH 7.4 and 37 "C
(physiological conditions) and the mixture is subsequently
analyzed. The progress of the Maillard reaction occurring on
proteins under such conditions can be followed fluorometrically.
2.3.2. Studies on Model Systems
For the analysis, the aqueous reaction solution or the reaction melt dissolved in water is repeatedly extracted with an
Angew. Chem. (02 (IY901 565-594
2.4. Initial Phase
The schemes shown are based on glucose as example for
the degradation of hexoses; deviations exhibited by pentoses
or disaccharides will be indicated. For simplification, openchain structures will mostly be used. In fact, the sugars and
many of their products exist mainly in their cyclic half-acetal
form. To start with, the structures of some reactants and
primary products are illustrated (Scheme 1).
,OH
R'
R'
R' = R2 = OH: u-Glucose
R' = NHR, RZ = OH. o-Glucosylamine 1 (always in p-form)
R' = OH, R Z= NHR: 2-Amino-2-deoxy-~-glucose2 (Heyns-product)
R'
R ' = R2 = OH: P-u-Fructopyranose (also a-form and furanose)
R ' = NHR, R2 = OH: P-u-Fructosylamine (not isolable with aliphatic
amines)
R' = OH, R2 = NHR: 1-Amino-1-deoxy-0-D-fructose
3 (Amddori product:
also a-form and furanose)
Scheme 1
2.4.1. Amines Reacting as Nucleophiles
The reaction of amines as nucleophiles leads, after addition to the carbonyl group of the sugar, to the formation of
567
giycosylamines. Relatively stable glycosylamines of type 1
are obtained from aromatic and heterocyclic amines, e.g.,
purines. A biologically very widely occurring substance of
this kind of structure is adenosine triphosphate. Glycosylamines of amino acids or aliphatic amines are usually only
detectable in very small amounts in sugar-amine reaction
mixtures. They quickly undergo further reactions and rearrange via the aminoenol5 into the aminoketose 3 and other
products.
Aminoketoses (or Amadori compounds) have been shown
to be present, alongside other products, in heated and in
stored
in dried fruit, in vegetable and milk product~:'~ as well as in soy sauce.'"] They also occur in infusion
solutions containing glucose and amino acids, intended for
parenteral nutrition (whether heat sterilized or mixed under
sterile conditions).[' They are also present in the human
body, in higher amounts in diabetics (cf. Section 4.4).The
biosynthesis of histidine proceeds via an aminoketose, which
is formed by rearrangement of an imine.
Deoxydiketoses and deoxyaldoketoses are formed as degradation products of the aminoketoses in the pH range 4-7.
1 -Deoxy-2,3-hexodiulose 6, 1-amino-l,4-dideoxy-2,3-hexodiulose 7, and 3-deoxy-2-hexosulose 8 are formed via the
-
HO HC - NH- R
HC=O
HC-OH
I
HO-CH
*NtYR
d
I
HCI
yc-
HF-
OH
-
tic- OH
H,C-
I
OH
H,C-
H2C- NH-R
OH
HC-NU-R
!I
I
c= 0
-
c-OH
I
A
HC - OH
7
2
I
1
HC -OH
HC-OH
HC-OH
H2C-OH
H,C-OH
H2C-OH
L
I
5
3
NH2
a : R = (CH,)~-CH
I
COOH
b :R=
3
c=o
I
c=o
I
HC-OH
1
HC- OH
I
H,C- OH
6
I
c=o
I
c=o
I
fH*
N H u
: R = (CH,)(-CH
I
I
coy
HC=O
I
I
FH2
HC-OH
H,C-OH
HC- OH
I
H,C-OH
10
11
Besides the aminoketoses 3, compounds of the type 12[13]
and 13114'are also obtained. The latter type of compound is
also known as an intermediate in the formation of osazones.
H,C
R
-N-
CH,
H,C-NH-R
I
c=o
C=O
HO-CH
C=N-R
I
HO- CH
HO-CH
I
HC-OH
HC-OH
HC- OH
HC-OH
HC-OH
HC- OH
H,C-
I
OH
on
H,C-
OH
H,C-
12
13
Ketoses react with amines via imines as intermediates to
form aminoaldoses such as 2, the so-called Heyns compounds." 51 In Nature, part of the biosynthesis of N-acetylglucosamine and N-acetylgalactosamine follows the same
principle: fructose reacts with ammonia and rearranges to
the activated glucosamine or galactosamine via the Schiff s
base. Compounds of type 2 are not very stable; they readily
n,con
I
- OH
HC
,
c=o
+
I
HO- CH
HC-
-
NH,-R
1
HO-CH
d
on
H,C-
I
HO-C -NH- R
HC-OH
--
I
no-CH
I
y-ou
HC-OH
H,C-OH
-
I
HC-OH
I
HC
,
-OH
OH
HC- NH-R
2
HC-OH
H,CHC=O
C- NH-R
HO-CH
I
HC-OH
HC-OH
H,C-OH
HC - OH
OH
C=N-R
I
HO-CH
I
HC- OH
HC- OH
H,C-OH
2
8
enediol4 or the aminoenol5. To demonstrate the conversion
of 3 into these products, trapping with o-phenylenediamine
is useful, the corresponding quinoxalines 9, 10, and 1I being
formed.'12] (The Group R in the formulas for the quinoxalines 9 and 11 is intended to indicate that, when Amadori
compounds derived from disaccharides react further, the
568
9
A
c= 0
I
OH
7
C
1
1
HC-OH
HF- OH
H,C-
CH-CH(CH,I,
coy
H,C-NH-R
y
I
H,C-OH
HO-CH
tic-on
HC-OH
I
H,C-OH
C-OH
I
HO-CH
I
I
1
HC-OH
11
II
I
I
1
c-on
HC-OH
Ho-w
HC-OH
I
H2C- NH-R
I Z
HC-OR'
I ,
HC-OH
HC-OH
A
F
OH
a:qH
a;IyR
c- OR'
1
OH
HC-OH
a"X?
HC=N-R
I
HCI
H0-m
OH
glycosidic bond in the corresponiinz ?ccx:izrores i s n o t
severed and therefore the quinoxalines bcar an a-glucosyl or
P-galactosyl residue, depending on whether they are derived
from maltose or lactose.)
react further, forming products such as the Amadori compounds.['6] Recent work suggests that aminoaldoses are also
formed in the human body. One can assume that Amadori
and Heyns compounds decompose in a parallel manner. In
presence of o-phenylenediamine, aminoaldoses yield predominantly the quinoxaline 11 via 8 as
In
view of the in vivo formation of Heyns compounds, further
studies of their properties are called for (cf. Section 4.2).
Angew. Chem. In[. Ed. Engl 29 (1990) 565-594
2.4.2. Amines as Bases or Acids
As bases or acids (depending on the prevailing pH), the
amines act as catalysts, for example, for the migration of the
carbonyl group of the sugar, i.e., aldoses are converted into
ketoses (fructose) and vice versa or an imine becomes an
aminoketose or aminoaldose (cf. Section 2.4.1) (corresponding to the Lobry de Bruyn-Alberda van Ekenstein rearrangement). Through loss of water, the enediols 14 and 15 give the
3-deoxy-2-hexosulose 8, and the l-deoxy-2,3-hexodiulose 6
and the 4-deoxy-2,3-hexodiulose 16, respectively. On heating
H2C- OH
HC-OH
HC=O
I
HC- OH
I
HO-CH
I
C- OH
-
HO-CH
3
HC- OH
I
HC-OH
c=o
-
I
HO-CH
7
H$- OH
HC-OH
I
OH
HC- OH
H2C- OH
H2C- OH
HC-
I
I
I
H,C
I
- OH
li
H,C-
OH
8
c=o
c=o
II
6
--
.
c-
OH
I
HC-OH
HC - OH
H2C- OH
15
on
H,C-
C- OH
-
80
8b
1L
11
Hone
H2F
0
"&OH
OH
c=o
H
OH
8c
8d
uct derived from these compounds is 5-hydroxymethyl-2furfural 18.["] In the presence of amines, the formation of 18
may be limited in favor of nitrogen-containing compounds.
The formation of metasaccharinic acid lactone 19 is noteworthy. Under alkaline conditions, the formation of metasdccharinic acid from 8 takes place by hydride migration.[191
In the Maillard reaction at pH 4-7, it is likely that the
lactone 19 is formed from the cyclic precursor 8 a through
Formation of the furaisomerization of the keto
none 20 from 8 a is conceivable after enolization involving
C3, followed by elimination of the hydroxy group from the
I
4
!HZ
HC-OH
H,C
HC= 0
I
-OH
c=o
16
I
CH
8 the sugar in the absence of amines, these reactions occur at
pH values below 3 f ' 8 1 or above 8 91 or on caramelization,
i.e., at temperatures above 130 "C. Thus, the significance of
amines in the formation of the deoxyosones 6-8 and 16 lies
in their ability to catalyze these sugar rearrangements under
pH conditions normally encountered in foods and living organisms (pH 4-7).
FH
11
HCH,C-
8a
P
OH
OH
17
\
Jn.
H2C 0
HO OH
HO
HO OH
CH
0
18
'L
2.5. Deoxyaldoketoses and Deoxydiketoses
As far as is known at present, many important reactions of
sugars during the Maillard reaction take place via deoxyaldoketoses or deoxydiketoses. These compounds can be
described as intramolecular disproportionation products of
sugars which undergo further reactions markedly more
rapidly than the original reactants; enolization, cyclization,
and loss of water thereby play major roles. The presence of
amines leads in addition to the formation of nitrogen-containing compounds. Products that are formed by cleavage of
the original sugar's carbon chain will be considered in a
separate section.
I1
HO OH
HO
19
OH
20
half-acetal function, but this has not yet been demonstrated.
It is known that 5,6-dimethoxyg1ucase 21 rearranges into
the corresponding compound 22.f241
The structurally equivalent aminofuran 24 is obtained as the main product when
HC=O
I
2.5.1. Reactions of 3-Deoxyaldoketoses
HC-OH
I
HO - CH
_
I
3-Deoxyhexosuloses were first shown to be present
amongst the reaction products of sugars and amines about
30 years ago.f2o1In solution, one would expect an equilibrium between the cyclic forms 8 a to 8d. The best known prod. 4 n ~ r w Chrm I02 li990) 565-591
HC- OH
I
-
H,C-I
H,CO
'$
OCH,
HC- OCH,
H,C-OCH,
22
21
569
N-acetylglucosamine 23 is heated in the presence of amines.
Compound 24 has been known for some time and serves
as indicator for N-acetylglucosamine or N-acetylgalact o ~ a m i n e . [In
~ ~Nature,
]
these aminosugars occur almost exclusively bound glycosidically to protein and so are not
available as reactants for the Maillard reaction.
enzyme-linked immuno sorbent assay (ELISA) has been de~eloped.[~
Lysylpyrrolealdehyde
'~
25 a can be determined by
reverse-phase HPLC after alkaline or enzymatic hydrolysis.I3'1
The monosaccharide derivatives 25 and 26 are also obtained by reaction of maltose and lactose with primary
amines. This can be understood as a consequence of 17 being
an intermediate; for its formation, loss of the oxygen atom
HC=O
attached to C4 is required, regardless of whether a glucosyl
I
HC- NHCOCH,
or galactosyl residue is attached to it, as in maltose and
HO-CH
lactose, respectively. Pentoses also react via the 3-deoxyalI
H,C-HC .I-!fNHCoCH3
0
HC-OH
doketoses
to give the corresponding pyrroles 251321
and pyriI
I
I
OH OH
HC- OH
diniumbetaines 26L271
(here, the hydroxymethyl group in
H
C
, - OH
2L
structures 25 and 26 needs to be replaced by hydrogen).
When reducing sugars react with secondary amines, hy23
droxymethylfurfural 18 is usually not formed or, if so, only
in small amounts. When the secondary amino acid proline is
In the Maillard reaction of the deoxyhexosuloses 8 in the
heated with compound 8, the so-called maltoxazine 31 can
presence of larger amounts of primary amines, the formation
of hydroxymethylfurfura118 can be largely suppressed. In its
place, are found pyrrolealdehydes of structure 25,[261as well
as pyridiniumbetaines of type 26.U2'l The fact, that 25 and 26
are obtained only in very poor yield from hydroxymethylfurfural and amines, indicates that the incorporation of amine
31
takes place in a precursor of 18. Further nitrogen-containing
products from compound 8 are the bispyrroles 27 and 28.'281
The structures allow one to assume a dimerization of the
be isolated as the main product. Maltoxazine is a volatile
component of malt and of beer and its formation will be
discussed in Section 2.8.2.[331The reaction of secondary
amines and pentoses leads to the formation of the orangeH,C
N CH
colored condensation product 35, which can be extracted
Hb
HO
with organic solvents as the main colored compound.[341The
25 0 : R = (CH,16-CH-NH,
26
formation of the hydroxyfuranone 34 will be considered latCO,H
er. Compound 35 is obtained in presence of 34 and secondary amines from the 3-deoxy-2-pentosulose 32 in good
b : R = (CH,)'- FH-NHyields
and from the cyclopentenedione 33 in very good
coyields. Compounds 31 and 35 are examples of carbocyclic
HC N CH2
CH
- -
1
!
n
y
HC
Iy3CH, _ 0 _H,C Jn
y $H
I;
0
HO
0
28
27
pyrrolealdehyde 25, whose hydroxymethyl OH group is subject to ready nucleophilic substitution (cf. Section 4.5). In the
presence of amino acids, additional compounds are found:
lactones of the type 29[291and lactams of structure 30.[221
For
the formation of these bicyclic compounds, the hydroxy
0
0
29
30
group of the pyrrolealdehyde 25 is again replaced by
other nucleophiles. In proteins, nitrogen-containing products are mainly linked to the peptide chain through the
&-aminogroups of lysine. To detect pyrrolealdehyde 25 b, an
570
HC=O
I
c= 0
HF-OH
H,&OH
3.4
0
+
NHI
0
p3
N-
33
35
32
products obtained from 3-deoxyaldoketoses with secondary
amines. They illustrate a general principle, found repeatedly;
namely, that sugars in the presence of secondary amines are
transformed into carbocyclic compounds, sometimes in substantial amounts.
3-Deoxyaldoketoses participate in the formation of yet
other pyrrolealdehydes. Thus, the compounds 37 and 38 are
formed upon reaction of 8 with the aminoketose 3. In both
cases, 36 functions as intermediate. The electron-attraction
by the formyl group in the intermediate 36 leads to retroaldol cleavage and thus to formation of compound 37.r351
In
a competing reaction, the enamine function induces formaAngex Chem. Inf. Ed. Engl. 29 (1990)
565-594
tion of the pyrrole 38.[36]The formulas refer to products
from hexoses, but the condensation proceeds analogously
for p e n t o ~ e s As
. ~ in
~ ~the
~ pyrrolealdehydes of type 25, the
hydroxy groups in the neighborhood of the pyrrole ring in 37
and 38 are readily substituted. Examples of this are the in-
I
I
H
I
H
H
43
H
Hy-OH
H~-OH
H2C- OH
L6
L5
HC-OH
2.5.2. Degradation of Z-Deoxydiketoses
38
1-Deoxydiketoses were shown only recently to be secondary products of aminoketoses.[12a * b1 They are converted
into compounds with the general reductone structure 47 (cf.
Section 3.4). At present, nothing is known about whether the
open-chain reductone 6 a is present in the reaction mixture.
One would expect an equilibrium mixture between the cyclic
forms 6 b to 6d in solution.
e
1
C- OH
C-Ctl
I
c=o
;-x
II
In addition to 3-deoxyaldoketoses, other a-dicarbonyl
compounds can also interact with aminoketoses. Thus, the
structure of the pyrrole 42 implies that 1,4-dideoxydiketose
41 participates in its formationr3' b1 (cf. Section 3.4). Further
evidence of similar condensation reactions comes from the
- -
L7
H2C
H o q0
:H
I
CH,
HO
A
c
CCH,
0
6b %
HC-OH
H,C-OH
6=
H
o
6a
c-Y
H2C-NH-
.--
II
i
LO
39
3
c=o
i
H,C-OH
H,C- OH
*
o=c
H
H,C-OH
6
C=O
H
HC-on
I
HW-
I
I
R=O
H
LL
y
cI = o
H
HC-OH
tramolecular ether formation in 39 and the partial replacement of the hydroxy group by a methoxy group during chromatographic purification (methanol/ethyl acetate as eluent)
of such compounds, leading to formation of, e.g., 40.[37bJ
3
I
H,C-OH
37
cI = o
H
tion of other primary products of the Maillard reaction, such
as the aminoaldoses, is provided by the structure of the
pyrrolealdehydes obtained on periodate oxidation. This is of
particular interest in relation to the Maillard reaction in
vivo, since here reactions following formation of pyrroles,
such as substitution of the a-hydroxy group, can lead to
cross-linking and other changes in protein structure (cf. Sections 4.5 and 5).
W=O
y
H
e
OH
CH,
6d
X = Y =O-,NI
Enolization and loss of water from the hemiketal form 6 b
of the I-deoxydiketose leads to the furanone 48J3*1 which
can be detected gas chromatographically after derivatization
(e.g., silylation). It is worth noting that furanone 48 isomerizes to compound 49 (R = H) only to an insignificant extent
-
HO-CH
I
7%
HC- OH
H,C -OH
H,C - OH
HzC - OH
I
Ll
42
structures of the pyrrolealdehydes, which are obtained from
the polar fractions from reaction mixtures of glucose and
pentoses with alkylamines after addition of sodium periodate. Apart from the pyrroledialdehydes and pyrroletrialdehydes 43 and 44, their isomers 45 and 46, respectively, are
also obtained, amongst other products, from 37 and
38.137 b. el
The isolation and identification of the pyrroles 37,38, and
42 proves that the aminoketoses are sufficiently reactive to
undergo condensation reactions. Evidence for the participaAngew. Clrem. 102 (1990) 565-594
OH - H2C=0
6
6 b - -
34
Hi)
HO
49
under the conditions of the Maillard reaction. Instead, formaldehyde is split off with formation of the pentose derivative
34 in considerable amounts. In this connection, the structure
of the colored main product 50 from the reaction mixtures of
hexoses and secondary amines is of interest, since it contains
57 1
H 0O NP : :
ducing, for example, C4-aminoreductones[461(cf. Section
2.6.1). Acetylformoin 62, which is formed from ketone 61 by
elimination of the P-hydroxy group attached to C6 of the
401readily undergoes further reactions and
original
is therefore only found, if at all, in low concentration in
model systems or foodstuffs.
50
the furanone 34 as a building block (cf. Section 3.3).[341
Furanone 48 can only be formed from disaccharides if the
glycosidic bond is broken. If one assumes that the glycosidic
bond is stilI present in the intermediate 6 b, then the glycosylfuranones 49 (R = a-glu, B-gal) should be formed. Evidence
for such compounds is, however, still lacking.
From pentoses and 6-deoxyhexoses, such as rhamnose,
the corresponding furanones 3413'] and 51 1401 are formed via
6
=
6d
11
58
11
59
i
0
COOH
the 1-deoxydiketoses. Compound 34 plays a crucial role in
the browning of pentoses (see Section 3.3). In the presence of
sulfur-containing compounds (e.g., cysteine), from which
hydrogen sulfide is liberated on heating, 34 also exerts a
considerable effect on aroma production. Substances are
formed that possess interesting aroma notes (mainly resembing heated meat;[411compounds 52 to 57 are given as exam-
52
53
54
55
56
57
ples). The corresponding S-containing products with an additional methyl group are obtained from furanone 51 on
heating with hydrogen sulfide. Compound 51 has a very low
odor threshold and an aroma note (fruity-roast-caramellike) that is perceived as pleasant. It is now manufactured on
a relatively large scale and is added to many
It
is interesting that 51 is formed also in plants.1431
Further products are obtained from 1-deoxydiketoses via
the cyclic hemiacetal form 6d. The pyranone 59 is formed by
enolization and loss of the hydroxy group from C2.L441
This
compound can be used as the universal indicator substance
for the occurrence of the Maillard reaction, since hexoses
occur in virtually all foodstuffs. Compound 59 is formed in
ppm amounts upon heating foodstuffs and is readily detectable by gas chromatography after extraction with an organic
solvent.14s1If the hemiacetal 6d loses the hydroxy group
from C5, the unstable 0-pyranone 58 is formed, and this
rearranges in part into the lactic acid ester 6OfZ3](cf. Section
2.6.1). Tautomerization of 6d leads to a P-diketone, which is
shown in its cyclic hemiacetal form 61. This compound has
not yet been shown to be an intermediate, but one can assume that it would readily undergo p-diketo scission, pro572
60
61
62
The reactions of 1-deoxydiketones derived from disaccharides differ in some respects from those of monosaccharide derivatives. This is understandable, if one starts with the
cyclic form 64.After enolization without splitting the glycosidic bond, only elimination of the hydroxy group on C5 is
possible, with formation of the P-pyranone 63. The y-pyranone 59 is therefore a product typical of monosaccharides,
which can only be formed via a free hydroxy group at C4.
The presence of small amounts of pyranone 59 among disaccharide reaction products can be explained by some degree
of hydrolysis of the glycosidic link during the Maillard reaction (cf. Section 2.5.1), the monosaccharide formed giving
rise to 59. The p-pyranone 63, which has so far been found
in heated milk (R = p-gal)[471and in steamed ginseng root
(red ginseng, R = ~ - g l u ) , [ ~is* a] key intermediate in disaccharide degradation. Its fate depends on the nature of the
glycosidically bound sugar. A B-galactosyl residue allows
formation both of five- and of six-membered rings. p-Galactosylisomaltol66 (R = p-gal), a Maillard product characteristic for lactose,[491is formed preponderantly. The corre-
6
li
OR
OR
: \\
64
1
66
67
CH,OH
1
65
Angew. Chem. In!. Ed. Engl. 29 (1990) 565-594
sponding reaction sequence can be observed with monosaccharides, when they are heated in alcoholic solution (in
methanol in the scheme). The y-pyranone 59 is converted at
pH 4 into the P-pyranone 67, which, on further lowering of
the pH to split the ketal, rearranges into the isomaltol methyl
ether66 (R = methyl).1471If these reactions arecarried out in
aqueous solution, isomaltol 66 (R = H) is formed only in
insignificant amounts,[501even though it corresponds to the
above methyl ether. Maltose leads to a higher proportion of
maltol 65,''
a-glucosylisomaltol 66 (R = a-glu) being
formed only in trace amounts.1521It follows that a-glucosylP-pyranone 63 (R = a-glu) must be subject to splitting of its
glycosidic bond to a considerable extent. This can be attributed to the greater steric hindrance to flexibility of the ring
of the maltose intermediate, which is not relieved by ringcontraction and formation of 66 (R = a-glu).
As already indicated, compound 62 is very reactive. The
so-called aminohexosereductones 68a are considerably more
stable. They are formed from 62 in presence of secondary
a m i n e ~ . [Under
~ ~ ] optimal conditions, aminohexosereductones can be obtained in yields of 3O%, based on the amount
of sugar employed. Indirectly, this shows the importance of
62 as an intermediate in the Maillard reaction.
The pyrrolinones 69a have been found in reactions between hexoses and primary amines. These compounds,
which can be obtained from 62 in very good
fluoresce strongly and, like the aminohexosereductones, possess
antioxidant properties (cf. Section 3.4).
Reactions of disaccharides with primary or secondary
amines also lead to aminoreductones. The galactosyl and
glucosyl residues remain glycosidically linked in compounds
68 b and 69 b, respectively, and the basic pyrrolinone and
cyclopentenone systems are identical with those of the
monosaccharide
5]
R- No a : H
I
CH,
68 a :
68 b :
R
~
69 Q : R = H
69 b : R = d-Glu, R-Gal
H
R =+Gluj13-Gal
The main product, extractable with organic solvents from
reaction mixtures of disaccharides and primary aliphatic
amines, is the pyridone 71.[s61Most likely 63 is first converted into the pyridiniumbetaines 70 (R = a-glu, P-gal), which
OR
63
0
-70
71
have recently been isolated in crystalline form from mixtures
of disaccharides and primary amine~.'~']
Pyridiniumbetaines
are predominantly converted into the pyridones 71 on cleavage of the glycosidic bond. Pyridones with this type of structure are known to bind metals such as iron[''] and aluminum
tightly;'591 a study of their effect o n the absorption and
A n p w Chrm. 102 11990) 565-594
elimination of trace elements is currently being carried out.
The pyrrole derivative 72, corresponding to galactosylisomalt0166 (R = P-gal), has been isolated as a byproduct from
reactions involving la~tose.[~'IUpon heating 66 together
with primary and secondary amines one obtains, inter aha,
the furan derivatives 73 and 74.C6O1Compounds 75[611and 76
I
73
72
I
71
75
R=H,CH,
are nitrogen-containing derivatives from 1-deoxydiketoses
of pentoses and 6-deoxyhexoses, respectively. So far, the
pyrrolinone 76 has only been obtained from reaction mixtures of condensation products of the type 77 and primary
amines. With secondary amines, compounds 77 rearranges
into compounds 78, the structures of which parallel those of
aminohexosereductones 68
-
O
m
o
H
c O
m o H
N CH,
F?C 0 CH,
I
R
- uHoU&
-N
77
76
HCR'
I
78
2.5.3. Reaction of 4-Deoxydiketoses
The formation of 4-deoxydiketoses 16 in Maillard reaction mixtures has so far not been confirmed. However, since
compounds have been isolated whose formation from sugars
is best explained in terms of 4-deoxydiketoses as intermediates, one may assume that such a-dicarbonyl compounds are
indeed formed.
Supporting evidence for their formation comes from the
structures of compounds 79-81. I n the presence of larger
amounts of primary amines, the formation of the hydroxyacetylfuran 79 can be largely suppressed in favor of the two
nitrogenous heterocycles 80 and 81.[26.
2 7 , 631
- -
0 OH
79
x - 0
80
X
2
81
N-
As shown in Section 2.4.2, 4-deoxydiketoses are not
formed from aminoketoses of type 3. The compounds 79-81
are therefore indicator substances showing that the reactions
have not proceeded exclusively via aminoketoses."
2.5.4. Reactions of I-Amino-1 ,I-dideoxydiketoses
Proof that 1-amino-l,4-dideoxydiketoses7 are reaction
products of the aminoketoses has only been obtained recent573
ly.['2b.c1Cyclization and loss of water leads to the aminoacetylfuran 82. As long as 30 years ago, the furan 82a was
obtained by heating Amadori compounds in the presence of
strong
(cf. Section 4.3). In the meantime, it has been
shown that 82 is also formed a t pH 4 to 7. It should be noted
that 82 is present as a stable salt under acid conditions, but
that above pH 5 it rapidly undergoes further reactions.
Among the rearrangement and condensation products
derived from 82 are the compounds 83 to 87."' 'I From the
structures of furoic acid 83, of its amide 84, and of furoylcarboxamide 85 it is immediately apparent that the aminoketone 82 is readily oxidized. The condensation to the
aminopyrrole 87 is noteworthy (compounds 86 and 87 were
isolated and identified as monoacetyl derivatives after acetylation of the total reaction mixture; the extent to which the
compounds exist in the free form as pyrroles has not yet been
ascertained). If such cyclizations are attributed to proteins,
then aminopyrrole formation by two &-aminogroups of peptide-bound lysine residues constitutes a cross-link. The significance of cross-linking reactions in biological materials
will be dealt with in Section 5.1.
In presence of ammonia aminoacetylfurans of type 82 give
good yields of the imidazole 88 o r its alkyl derivatives 89
respectively.[651Compound 88 has been described recently as
indicator substance for protein cross-linking (see Section
4.5). 88 was assumed to be derived from a cross-link. Howev-
has been allocated only on the basis of the mass spectrum of
a peak in the gas chromatogram of a silylated sugar-amine
reaction
2.6. Cleavage Reactions
It has already been mentioned several times that, during
the reactions undergone by sugars, cleavage of the carbon
chain can occur. As will be shown, some cleavage products
are very reactive and can substantially accelerate the Maillard reaction.[671Sugar fragmentation products play a considerable role in aroma formation (cf. Section 3.1).
2.6. I . Retroaldolization
Sugars and many of their derivatives can undergo retroaldol-type reactions. The structures of compounds 92 to
102[46a. 681permit one to assume that they have been formed
HCHO
HOH,C-CHO
93
92
HOH,C-
OHC-CHO
CO -CH,OH
~
0
99
98
HOH,C-CHQH
101 Q:X=OH,Y=N101 b :X=N-,Y =OH
0
NHR
96
H3C- CO-CO-CHI
H3C- CO- CX= CHY
100
H3C-CO-CHC
95
94
HOH,C-CO-CHO
97
H3C- CO-CO-CH,OH
f
H,C-CO-CH20H
-CHOH-CHO
102
I
NH2
82 a : R = c(H$-,I,
83
R'= OH
84
R'=NHR
85
R':
/
7
CNHR
II
0
HC-NHR
N
86
CH
CH
N
o/
X =OH
R
O
88
R = H
89
R =Alkyl
by a retroaldol cleavage. In some cases, oxidation and/or
dehydration follows. Some of the cleavage products are very
reactive, readily undergoing condensation reactions. The
available partners consist of other retroaldol compounds, as
well as of sugar derivatives with carbon chains still intact.
The condensation products 103 to 11 1 illustrate the type of
structures formed.r69,37 Further research has shown that
the progress of the degradation of the sugar is paralleled by
the formation of the cleavage products. In the early phase,
I
90
91
er, it became evident[651that the formation of 88 takes place
from furosine 82a and ammonia during and subsequently to
the acid hydrolysis of proteins. 82a is itself formed in the
process, as already mentioned, from the lysine Amadori
compound 3a or 3c, and ammonia is liberated from asparagine and glutamine.
Recently, the isolation of aminoreductone 90 from a glucose-propylamine reaction mixture was achieved.[37dl It can
be assumed that a 1-propylamino-l,4-dideoxydiketose
of the
general type 7 is formed as intermediate. This shows for the
first time that an open-chain C,-aminoreductone is present
in a Maillard reaction mixture. The formation of the furanone 91 can also be derived from 7, but so far structure 91
574
@:-
HO
103 a : x = r ; R = OH
103 b : X = O , R = O H
104 a : R=CHO,R'=H
105 a : R=CHO,R'=H
1OL b: R=COCH,R'=H
103 C X = Q,R = NH-
104 C : R
105 b : R=CHO,R'=CH,
105 C R = COCH,,R'=H
COCH$=CH,
-HNOC-CH-
106
107
R=H,CH,
R=H,CH,
108
CH,
109
110
1110
lllb
Angew. Chem. In!. Ed. Engl. 29 (1990) 565-594
C, fragments preponderate, which have been formed from
the sugar or the imine 1 provided that an aldose had been the
starting point. Electron spin resonance spectra have provided evidence for the partial conversion of the imine cleavage
product 112 into the pyridinium radical 113, from which
oxidation produces glyoxal or the corresponding nitrogenous analogs that immediately react further.[701
amines), but with exclusion of oxygen, extraction with organic solvents yields amongst other products the lactone 118,
formic acid 119, and furfuryl alcohol 120.r741
Model experiments with glyoxal and methylglyoxal had previously al-
-+
8
NH,-
HC
,
OH
HO
HC - NHR
1 - -
I1
HC- OH
+
112
R
--@
[) - R
118
HC=X
I
X = Y 10,NR
I
H F -OH
Somewhat later, C, compounds appear, but by then ketoses or Amadori products are already present. The amounts
of the cleavage products found depends on the pH of the
reaction medium, including the foodstuff. At pHs below 7,
retroaldol products are likely to play only a subordinate
role,[67.711
but one must remember that even very small
amounts of aroma compounds formed in this way may have
great impact, if their odor threshold is very low.
If one assumes that the rate of browning depends on the
amounts of the C, and C, products formed, it would follow
that solutions of disaccharides become colored more slowly.
Certainly, the C , products can be formed by retroaldolization only after hydrolysis of the glycosidic
Possibly
the configuration of the hydroxy group at C4 of glucose and
galactose also affects the rates of cleavage and of browning.[731
Labeling experiments have shown that C, aminoreductones 101 are formed via the P-diketone 61. Neither C1 nor
C6 of the sugar gives rise to the methyl group of the C,
product 101.[461
After cleavage of 61 to the tetrose (shown
here in the enolized form 114), the deoxy form 100 incorpo-
61
OH
-
0
119
120
HC= Y
113
HC=O
m
HO
ready led to the identification of the acid amides 121 a and
121 b.[751The formation of 118-121 requires cleavage of the
a-diketo group. Such cleavages are known, even though
there are hardly any proposals as to how they occur.
--
RC=Q
+NH,-
I
RC= 0
R = H,CH,
RC= 0
I
NH-
121 O : R = H
121 b : R = CH,
For the decarboxylation of the a-ammoniumcarboxylic
acid 122, an immoniumbetaine of the type 123 has been
indicated as an
which has some resem-HF-COOH
+
NH
I
-
-
-C -COOH
I1
*CH
122
-C?
1 1 - I
-c:
YH
'NH
123
blance to the thiazoliumbetaine 124 derived from thiamine.
If one assumes that, in the fission of the deoxyosone 8, imine
125 is formed first, then it is to be expected that alongside 118
C,-immoniumbetaines will be formed. However, so far
- - 100 - - 101
114
rates amine to give the product 101. Retroaldolization is also
involved in the formation of the acids 116 and 117: the
unstable p-pyranone 58 is in equilibrium with the furanone
115, which on P-cleavage gives the ester 60. Depending on
the conditions, this, in turn, is split more or less rapidly into
P-hydroxypropionic acid 116 and lactic acid 117.1231
FOOH
0
115
116
117
8
I
C=N-
- - 118
H H
125
proof or identification of such derivatives has not been unequivocally achieved. Alternatively, the intermediate 126
could lead to formic acid 119 and the C, fragment 127, the
latter reacting further to give furfuryl alcohol 120. Lactone
118 and formic acid are formed in considerably greater
amounts in the presence of oxygen, whereas the amount of
furfuryl alcohol 120 decreases. It would seem therefore that
lo
8
Angcic. Chrm. 102 11990) 565-594
H,CAQ?":
HO
2.6.2. Cleavage of a-Diketones
If the 3-deoxyaldoketose 8 is heated under Maillard conditions (i.e., in neutral aqueous solution in the presence of
--
- -"'ON;
- 119
H,C
I
HQ
0
H
HOT$
+ HZC OH
- - 120
HO
126
127
575
118 and 119 are formed in the reaction mixture predominantly by oxidative degradation of 8.
Recently,[771 it proved possible to identify carboxymethyllysine 128 in the protein hydrolyzate of biological
material (cf. Section 4.5). From the experimental results ob-
13-4 x = o
135 x = NH
136
pyridine 136.[831Compounds such as, inter aha, 137, 138,
and 139 are derivable from I-deoxydiketoses of type 6. However, there is evidence that in these compounds the methyl
group is not derived from Cl of the hexose. Therefore, a
degradation via the imine 1 a has been proposed as a route to
the formation of 138 and 139:[841
loss of water from 1a leads
128
tained to date, it follows that the formation of 128 is due to
oxidative cleavage of the corresponding aminoketose 3
137 x :O
138 X = N H
2.7. The Strecker Degradation
2.7.1. Strecker Degradation of Amino Acids
The decarboxylation of amino acids on heating with sugars has been known for many years. The decarboxylation is
actually brought about by reaction of the amino acids with
a-dicarbonyl compounds.[791The azavinylogous P-keto acid
129 is formed first and then loses carbon dioxide. The significance of the Strecker degradation lies in the fact that the
amino acids furnish ammonia and the reactive aldehydes of
type 131, compounds that can undergo, inter alia, condensa-
first to the 2-deoxyhexose 140, which, after hydrolysis of the
imine function, reacts further to yield the pyrrole 138 o r the
pyridine 139. The Strecker degradation of proline and hydroxyproline will be considered in detail in Section 2.8.2.
Another reaction sequence, that also leads to the reduction
of a-dicarbonyl compounds, will be dealt with in Section 3.4.
HC=N-CH-COOH
I
I
HC- OH
HO-CH
-
I
c=
0
I
+
H,N-CH-COOH
I
I
cI = o
R
1
R
129i
I
C=O
I
H C - OH
I
+
NH,
-
I
C=N-CH-COOH
'
c
c
I
I
--
139
I
140
la
I
2.7.2. Strecker Degradation of Amines?
C-NH,
C-OH
I
O= CH
130
131
R
tion reactions. Thus, in reaction mixtures including lysine
and phenylalanine as well as hexoses, furans 132'801and
133[8* have been found, whose structures comprise the corresponding Strecker aldehyde. Ammonia in the free form o r
as the aminoenol 130 participates in the formation of
pyridine, pyrazine, imidazole, and other
In many foodstuffs and especially in vivo, the lysine
side chain is the most important amine for the Maillard
reaction. This raises the question of whether, in the presence
of a-dicarbonyl compounds, the formyl derivative 141 is produced. For proteins it would seem likely that reactions would
follow that would be comparable to those that take place
during the synthesis of collagen and that lead t o cross-link-
NH,
The Strecker degradation of amino acids brings about the
reduction of the deoxyaldoketoses and diketoses. Cyclization and loss of water converts 8 into the furan 134 and, with
the participation of ammonia, into the pyrrole 135 and the
576
II
HC
C-N=CH
II
I
C-OH R
i
I
HC- N= CH
-
HO-CH
I
C=O
139
cI = o
1L1
ing. It is known that primary aliphatic amines in presence of
vicinal triketones such as ninhydrin and dehydroascorbic
acid are converted into the corresponding aldehydes.[851In
the Maillard reaction, triketones, such as the J3-pyranone 58
and oxidized reductones, are indeed formed, but the a-diketo
compounds are of far greater significance. The experimental
results available at present d o not yet provide unequivocal
proof that a-diketo compounds can lead to the splitting of
Angex Chem hi.Ed Engl. 29 (1990) 865-894
primary aliphatic amines into ammonia and aldehydes under
the conditions of the Maillard reaction.
since been found in foodstuffs that have been particularly
severely exposed to heat.["]
2.8. The Influence of Particular Amino Acids and Amines
on the Course of the Reaction
2.8.2. Proline and Hydroxyproline
This section is devoted to some special reactions that, due
to the specific nature of the amino components, lead to unusual products.
2.8.1. Tryptophan
An important finding, particularly with regard to the
Maillard reaction in vivo, is that protein-bound tryptophan
reacts with reducing sugars only to a limited extent.[861Thus,
this essential amino acid is, in contrast to the essential lysine,
hardly damaged in protein-sugar reactions. In foodstuffs,
tryptophan can also occur in the free state (e.g., in soy sauce).
Then. even under mild conditions, reaction with carbonyl
compounds and ring-closure with formation of P-carboiines
become possible. The tetrahydro derivatives 142 and 143
have been shown to be present in soy sauce1871(cf. Section 3.5).
Under drastic reaction conditions, further P-carbolines
are formed, some of which have also been found in foodstuffs.["] It is apparent from structures 144-148 that,
alongside cyclization, decarboxylations and dehydrogenations take place. It is likely that 3-deoxyaldoketoses participate in the formation of compounds 144 to 148, which have
been isolated from pentose-tryptophan reaction mixtures. If
Proline occurs in relatively large amounts in cereals and
their products, such as malt. The participation of this imino
acid in the formation of volatile compounds contributing to
the aroma of malt has been particularly thoroughly researched. It is assumed that the dihydropyrrylium compound 152 plays a central role in the process. 152 is an
intermediate of the reaction of proline with a-dicarbonyl
compounds. The cation 152, which is obtained by decdrboxylation, leads to a series of nitrogen-containing hetero-
152
cycles by a variety of reactions. From the structures of compounds 31 and 153-162, it can be deduced that the stabilization of the intermediate 152 is achieved by electron uptake
(153), loss of protons (154), addition of heteroatoms (31), or
qcooH
@@
0
0'
R
W
144
142 R = H
1L3
R-CYOH
R
145
1.46
1L7
R=%
148
R=Q
R=H
R=COCY
the amino group is blocked, formation of glycosyl derivatives 149 (R = P-Glu, p-Xyl) has been observed. Evidence of
their further reaction, e.g., to aminoketoses via Amadori
rearrangement, has not been obtained so far.[s91To what
extent products of the type 149 are formed in the Maillard
reaction has not been investigated. More attention has been
devoted to the tryptophan derivatives 150 and 151, because
of their strong mutagenic and carcinogenic activity.[901
These substances, which were first detected in pyrolyzates of
tryptophan and of tryptophan-containing proteins, have
addition of other nucleophiles (155).[921
Opening of the dihydropyrrylium ring, followed by intra- or intermolecular condensation, leads to renewed ring closure, for example, with
formation of pyridine derivatives 156- ~ 9 ,as ~well~ as~of '
156
157
R' = H, CH,
R:R';H
R = H,OH
R
158
R = CH,%H,
159
R = H, CHI
H,@=CH,
R = Cy,d = H
H= 'RJ l&T=R
the dihydroindolizine 160, and the partially hydrogenated
cyclopentazepinones 161 - 162.1941 For the formation of
maltoxazine 31, a pathway starting from 8 via 152a has been
proposed.
H ydroxyproline occurs almost exclusively in connective
tissue protein (collagen) and therefore is of no great signifiAngeu. Ch<,rn.102 (1990) 565-594
517
161
162
R=H,CH,
R=R'=H
R=
H,R'= c q
R = OH,@= CH,
C-OH
I
y
--
2
HC-OH
I
H 2I C - \ 3
C=O
y
2
y
2
2.8.4. Arginine
7
0
1
c=o
HC-OH
I
I
H2C- OH
1
H2C-OH
31
152a
cance for the Maillard reaction in vivo or in foodstuffs.
From model systems, it appears that, as with the degradation
of proline, the dihydropyrrylium compound 163 leads to
volatile products. In the process, stabilization occurs preponderantly by loss of water and formation of a pyrrole ring
(compounds of the type 164), which may be built into a
bicyclic system (compound 165).Ig5]Hydroxyproline has a
i2Nq - $-w
OH
found, inter aha, in the volatile fraction of roasted coffee
beans[981and of cooked meat.[7j*991
Some important sulfur-containing aroma compounds are
dealt with in Section 3.1. Cysteine has a retarding effect on
the Maillard reaction, and this will be discussed in Section 3.7.
I
C - N ~
- - R-NZ
C -OH
I
163
16L
..
R = CH,COCH,,
CH2COCH,CH2COCH3,
b
0 C
H
2
0
H
165
'li0i(R',
In reports on the damage suffered by amino acids when
proteins are heated in the presence of sugars, reference is
frequently made to the fact that arginine, as well as lysine,
undergoes extensive modification during the Maillard reaction.['"' Changes to lysine are often tantamount to loss of
this amino acid and are therefore important for the nutritional value of the food, since lysine is an essential amino
acid (cf. Section 3.6). During the in vivo Maillard reaction,
lysine participates in cross-linking. Recently, it was discovered that arginine can also undergo such cross-linking reactions. From the structure of the product 167 it follows that
arginine and lysine are linked via a imidazopyridinium
bridge""] (cf. Sections 4.5 and 5.1).
It is worth noting that the Maillard reaction is not initiated
by the guanidino group.['"I Heating glucose with N"-acetylarginine causes no browning of the solution and the hexose
remains largely unchanged. The guanidino group of arginine
bound in a protein therefore needs reactive Maillard products in order to be converted into, e.g., 167. Heating aqueous
solutions of methylguanidine (as models for arginine) with CIand f3-dicarbonyl compounds at pH 7 to S led to the isolation
of, e.g., imidazolones and pyrimidines of the general type 168
and 169, respectively (the formation of the imidazolones re-
(R'= H,CH,,CH20Hl
certain significance, as does proline, as a starting material for
the formation of volatile components of tobacco smoke.
2.8.3. Cysteine and Cystine
Cysteine and cystine do not occur free to any large extent
in foodstuffs. The great significance of these amino acids for
aroma formation is mainly due to the fact that, on heating,
hydrogen sulfide is released. Hydrogen sulfide reacts with
Maillard products to give a multitude of sulfur-containing
compounds. Their important influence on the aroma is in
part due to their extremely low odor thresholds. It follows
that even small amounts of cysteine or cystine can affect the
aroma perceived. If the extensive patent literature and the
results of model experiments are taken as a guide, cysteine
and cystine play an important role in the formation of the
aroma of meat.[961From model systems of cysteine with
pentoses and hexoses, a broad spectrum of sulfur-containing
volatile compounds has been isolated and some of these have
very intense odors.r971Mono-, di-, and trisubstituted thiazoles of the general structure 166 predominate. Thiazoles are
quires at least one aldehyde function).['031Further research
is required in order to ascertain which Maillard products
with this function react preferentially with the guanidino
group.
k=f
tion has focused on these meat constituents, since they participate in the formation of the carcinogenic and strongly
syN166
578
2.8.5. Creatine
Creatine 170a and creatine phosphate 170b occur in relatively large amounts (0.5 %) in muscle. Heating causes mainly cyclization to creatinine 171. For some years now, atten-
170a R = H
170b : R = PO(OH),
H2C-CQOH
H3CN
II
do
H3C,NkNH
171
HN=C-NHR
Angew Chem. Int. Ed. Engl. 29 (i990) 565-594
mutagenic compounds 172 and 173. Both compounds have
been shown to be present in high-temperature roasted mixtures of amino acids and creatine
and without
h e x ~ s e s ~and
' ~ ~later
] also in meat and fish products.['061
&CH,-CH-CH-CH,
I
l
l
I=?,
N O H
OH OH OH
N\YNH
R
178
176 0
R = CH-CH-CH-CH,
I
l
l
OH OH OH bH
176 b : R
It can be assumed that creatinine, possessing a reactive
methylene group, is able to condense with aldehydes (e.g.,
formed by Strecker degradation of amino acids) and that
further reactions with sugar degradation products and ammonia follow. Certainly the concentration of 172 and 173 fell
when Strecker aldehydes were trapped with tryptophan.[1071
However, such a method cannot readily be adopted for foodstuffs, since tryptophan can form mutagenic and carcinogenic compounds with sugars and aldehydes (cf. Section
2.8.1).
Meanwhile, it has become apparent that the amount of
carcinogens formed is largely determined by the conditions
under which the foodstuff has been treated and that mild
processing can to a great extent avoid their production. For
example, if the food is heated to above 130 "C, the concentrations of 172 and 173 formed lie in the lower ppb range, but
cooking at the boil leaves them below the limits of detection.
= CO-CH,
latter in foodstuffs is not significant, as already mentioned in
Section 2.5.1.
2.9. High Molecular Weight Compounds (Melanoidins)
If the reaction mixtures of sugars and amines are submitted to exclusion chromatography, fractions can be obtained
with molecular weights of about 7000 Da and even greater.
So far it has not yet proved possible to isolate homogeneous
high molecular weight Maillard products." 161 Very little i s
known about the structure of the high molecular weight
substances, the so-called melanoidins. Knowledge has however been gained from the interpretation of spectral data,
although some of the 'H- and 13C-NMR spectra of the
melanoidins isolated by different research groups have differed substantially. Melanoidins with groups of signals that
largely resemble those of Amadori compounds" 71 contrast
with those that exhibit signals in the olefin, aromatic, and
carbonyl regions.["'] Experiments with 3C- and 5N-labeled sugars or amino acids have shown that the carboxy
group of the amino acid survives in part in the melanoidin
and that the signal of the a-carbon largely remains in the
position in the melanoidin spectrum where it appears in the
amino acid spectrum. Somewhat divergent from this, the
"N-NMR spectrum also exhibits signals in the pyrrole, indole, and amide regions." '] Signals for the Cl atom of the
sugar can be found in all parts of the spectrum, consistent
with the formation of different deoxyaldoketoses and diketoses and their derivatives. Results, obtained so far by the
degradation of melanoidins with ozone and hydrogen peroxide,["'] have not led to any illuminating conclusions as regards the monomers involved in the condensation to
melanoidins. Combined pyrolysis-gas chromatography of
melanoidins, even when combined with mass-spectrometry,
has not so far given unequivocal results either." I' Differently produced caramel colors have been differentiated by
means of Curie Point pyrolysis.['
Electron spin resonance
has also been applied to melanoidins, but so far no deductions have been drawn as regards the structure of the radicals.['221The absorption of melanoidins in the ultraviolet
and visible regions shows that condensation reactions have
participated only to a limited extent in the linking of the
monomers.[' 231 It is possible that substitution reactions play
a role, as has been discussed in relation to the cross-linking
of proteins.
'
'
2.8.6. Ammonia
Unless foods have been allowed to spoil, they normally
contain ammonia only in very low concentrations. To manufacture caramel as a coloring matter for foodstuffs (e.g., cola
beverages), ammonia in large amounts is brought into reaction with sugars. With ammonia, the formation of pyridines,['O'I pyrazine~,['~'land imidazoles" O1 is possible without requiring the cleavage of a C-N bond. As a consequence,
these heterocycles occur in relatively large amounts in sugarammonia reaction mixtures. Compounds 174- 177 are given
'
CH-CH-CH-CH,
/
I
l
l
I
17L R = CH2-CH-CH-CH,
I
l
175
l
l
OH OH OH
R
CH =CH- CH- CH,
I
l
OH OH OH
1
CH-CH-CH-CH,
AH &I &-I &I
I
OH OH
as examples. They can serve as evidence of the addition of
ammonia caramel.[' a, 'I This can be important for beer,
bread, and other foods, where dark color can simulate the
use of certain raw materiaIs.["21
Compound 176b is known to reduce reversibly the number of lymphocytes in rats"
when fed a diet low in vitamin B,. The methylimidazole 178 has toxic properties." 1 3 ]
For both substances, upper concentration limits have been
laid down for ammonia caramel." 14] Polyhydroxyalkylpyrazines are formed also from, for example, amino sugarS[1 151 (e.g., 2-aminoglucose), but the occurrence of the
''
'
3. Significance of the Maillard Reaction for Foods
When fire was introduced as an aid in the preparation of
foods several hundred thousand years ago, the Maillard re-
579
action became intimately linked to food chemistry. As evident from the recipes in cookery books down the ages, the
Maillard reaction serves in particular to enhance the sensory
qualities of many dishes. Who does not recognize the typical
aroma of warm, crusty bread, of roasted meat, or of freshly
ground roast coffee? The brown shades of color of different
beers, as well as of many foods that have been exposed to
heat, are equally the result of the Maillard reaction. However, for some years now, it has been known that the Maillard
reaction is also involved in the formation of mutagenic and
carcinogenic substances and can cause a reduction in the
nutritional value of foods.
3.1. Aroma
The great interest shown by the food industry in the Maillard reaction largely stems from a desire to produce and
control the characteristic aromas obtained on cooking, baking, roasting, and grilling. Once the analytical technique of
combined gas chromatography-mass spectrometry had
been developed for the separation and identification of relatively readily volatile substances, the search for compounds
with specific odors became greatly intensified. The results
have been summarized in a series of review articIes.['J. '241
Herein, therefore, we shall deal only with certain aspects.
Almost invariably, it has proved possible to isolate and
identify from each foodstuff (boiled, baked, roasted, etc.)
hundreds of volatile compounds. The structures of compounds 166 and 179-185 are intended to give an idea of the
I
179
1_1
SfN
166
180
181
182
?Ff
NYNH
x%
XX
183
18L
185
odor port. By diluting the mixture in a stepwise manner
before injection (usually by doubling the amount of solvent
at each step), progressively fewer substances remain to be
detected by the nose. The substances left in the later stages of
the process usually carry most of the aroma of the mixture.
For example, aromagrams have shown that it is not so much
the trithiolane 186 and the dithiazine 187, as previously
t h o ~ g h t , [ ' ~that
~ 1 are important for the aroma of boiled
meat, but the thiofuran 188 and its oxidation product
189.f'261
The acetylpyrroline 190 is one of the important
contributors to the aroma of white bread,['271 whereas the
typical odor of roasted nuts is in part determined by the
pyrazines 191 and 192.["*] Recently, it was shown that the
concentration of the acetylpyrroline 190 is substantially dependent on the amount of C , products in the bread. Particularly in white bread, yeast, as well as the Maillard reaction,
is an important source of these compounds (e.g., dihydroxyacetone m o n ~ p h o s p h a t e ) . [ ' ~ ~ ~
190
191
192
Depending on the detailed processing conditions used and
bearing in mind the variability of biological raw materials, it
is not surprising that the concentrations of the intensely
odorous components in foods differ greatly from sample to
sample. This makes for difficulties in regulation, surveillance, and specification. Proof of the addition of aroma compounds that could have been formed by the Maillard reaction is therefore hardly possible.
3.2. Taste
multiplicity of the heterocyclic systems encountered. To
these must be added the many components based on proline
and mentioned in Section 2.5.2, as well as other volatile
Maillard products, such as esters, acids, ketones, aldehydes,
alcohols, alkenes, alkanes, aromatic derivatives, amines, and
mercaptans, that contribute to the aroma positively o r negatively to a greater or lesser extent.
So far, all hopes have not been fulfilled; in particular, the
aromas of boiled or grilled meat, of roasted coffee, of chocolate, and of bread cannot yet be satisfactorily reproduced by
means of a single substance and one now suspects that this
will never be the case. In general, reproduction of such aromas requires several components, which need to be present
in the correct proportions. The so-called aromagrams allow
one to estimate the contributions of the individual components to the final aroma. Aromagrams are produced by assessing with the nose, at the exit of a gas chromatograph, the
volatiles separated from a complex mixture. Decisive for
detection of each compound is whether its concentration in
the mixture is sufficient for it to exceed the threshold a t the
580
When bread or meat is overheated on its surface during
baking or roasting, respectively, the crust usually tastes bitter. The same effect is noted when wort boiling takes place at
higher temperatures. Also model mixtures of sugars and
amino acids taste bitter when they have been heated under
grilling conditions. The most strongly bitter note in such
experiments has been achieved with proline." 301 From reaction mixtures of this imino acid and hexoses, bitter substances 193-195 were successfully isolated.[94c.1311 It is
noteworthy, that the taste threshold of the bispyrrolidine 193
lies higher by a factor of only 5 than that of quinine, i.e., 193
is already recognized as being bitter at concentrations of a
few ppm. Investigations have yet to be carried out on the
part played by compounds 193-195 in the bitter taste of
highly heated proline-containing cereal products. A bitter
taste is also exhibited by the proline-specific cyclopentazepinones 161 and 162r94b1
(cf. Section 2.8.2), as well as
by the pyrroles 196 and 197, which are formed on heating
serine and threonine with hexoses.[' 321 However, with some
Angew. Chem. Int. Ed. Engl. 29 (1990) 565-594
y
2
HC-OH
193
19L
195
R =OH, R‘=CH,
R=H,~=cH,
R = R‘= H
R
196 R = H
197 R = CH,
of these compounds, the taste thresholds, so far as they have
been investigated, exceed the concentrations in which they
are formed.
pyrrolinone 200 is obtained on heating furfural in the presence of a primary amine and that its structure is very like that
of the furanone 201.[’341One can deduce from this the great
potential for condensation which is inherent in the 3-deoxyaldoketoses, the precursors of these colored compounds.
In foodstuffs, the condensation could occur with a whole
range of carbonyl compounds, so that a multitude of browning products could result by means of this type of condensation alone, quite apart from whether 200 or 201 themselves
are stable end products. Indeed, this makes it understandable why the attempted isolation of colored Maillard products from heated foods has not yet been successful.
3.3. Color
From the fact that “we also eat with our eyes”, the significance of colored compounds in food chemistry can be deduced. However, the Maillard reaction contributes not only
to desirable formation of color, but also to discoloration of
foods. which is synonymous with a lowering of quality. The
determination of the degree of browning (usually through
absorbance at 420 nm) is often used analytically to assess the
extent to which the Maillard reaction has taken place. Since
the concentrations of the sugar and amino components in
foodstuffs are very variable, measurement of the color intensity does not readily yield comparable results.
The isolation and identification of colored Maillard products has so far been achieved only with model systems. The
structures of substances, in which the carbon chain of the
sugar has remained intact, show that the deoxyaldoketoses
and diketoses are important intermediates. Thus, the yellow
compound 198 is formed by condensation of hydroxymethylfurfural 18 with the pyranone 17a.[133]Both of these
compounds are derived from the 3-deoxyaldoketose 8. The
analogous product 199 is present in reaction mixtures con-
OH
Hb
+
Hi-cq:
HCR
HCR
200
201
I-Deoxydiketoses can also lead to colored substances. Interaction of Amadori compounds with carbonyl compounds
leads, inter alia, to the yellow compounds 202 and 203,“ 351
- -
6
I
+
59
62
I
I
I
1
R C n o I R’QH
1
CH =CHR
203
202
8 -
17
-18
--r---
RQH
1I
flQ2
H2C
Q
.l!k CH
Htf
Q
HQ
fl
n2y o
OR
HQ
198
OH
HO
17a
199
taining pentoses. These compounds may be stabilized as acetals in systems containing little water. Such systems include,
for example, bread crust, from which even a short time in the
baking oven removes most of the water. Reaction conditions
are then favorable for acetal or ketal formation with alcohols
(e.g., carbohydrates). In aqueous solution, the hemiacetal
198 (R = H) is not stable; the furanone 201 becomes the
main colored product when a carbonyl compound (e.g., furfural) is added to the solution. It can be assumed that rearrangement occurs first and that subsequent condensation
leads to the furanone 201.[1331
It is noteworthy that the red
Angrit
Chvm I02 (1990) 565 594
RCHQ
HCR
parts of which are clearly derived from I-deoxydiketoses via
59 and 62, respectively. The furanone 203 becomes stabilized
as the ketal in media poor in water; its hemiacetal quickly
reacts further.
Compound 50 (cf. Section 2.5.2) becomes the main colored product, when, to simulate media poor in water, heating is carried out in alcoholic solution. The mechanism
shown is only hypothetical. As already mentioned, the fwdnone 48 readily loses formaldehyde with formation of the
hydroxyfuranone 34.1 -Deoxydiketose reacts in alcoholic solution to some extent as the reductone ether, which rearranges by loss of water and incorporation of the secondary
amine to the cyclopentenone derivative 204. This is subject to
60
- 63 -
I
OR
11
,l+
20L
581
nucleophilic attack by the furanone 34 and oxidation then
yields the colored product 50.
Research on model systems has produced much data on
the browning of pentoses. The structures of the yellow main
products 77a and 77 b, which are formed on heating pentoses
with primary a m i n e ~ , suggest
~ ' ~ ~ ~condensation of the hydroxyfuranone 34 with furfural and pyrrolealdehyde, respectively. The formation of the colored product 35 from pentoses and secondary amines has already been mentioned in
Section 2.5.1.77a,77 b, and 35 also contain the hydroxyfura-
208, must be expected to be found o n the surface of smoked
meat and fish products.
There are further possibilities for color production during
the Maillard reaction. Thus, Maillard products which possess a carbonyl function next to an enol grouping often give
colored complexes with metal ions. Aqueous solutions of
ferric salts turn red, violet, or blue on addition to solutions
of compounds 34, 48, 50, 51, 59, 62, 65, 68 (R = H), 69
(R = H), 71, 90, 101, and 209.[1371
(Substance 209 is obtained from aminoketoses in the presence of oxidizing
agents.) The significance of such colored complexes in the
browning of heated or stored foodstuffs remains to be assessed.
77a: X = 0
7%: X NHO
H2C-OH
none 34 as part of their molecule. 34 is able to condense
under mild conditions even with less reactive aldehydes and
ketones to give colored compounds of type 77, and, with
reactive carbonyl compounds, it will even react further to
give products of structure 205.1621
Bearing in mind that, un-
34
RCOR'
R,omoH
RCOR'
@,C
0 CH,
77
A further possibility for the formation of colored products
lies in the reaction of a$-triketones and a-dicarbonyl compounds with amino acids and peptides. As shown in Section
2.7, the Strecker degradation produces aminoenols of type
130, which can react with further a-diketones or u,&triketones to form colored condensation products of the general
structure 210.[1381
Peptides give compounds of type 211.[1391
k=o
I
c=o
$,RC ? X 0o Hc=c,
,R
+
- ,,
I
1
-H2N-CH-C0,H
I
C-OH
o=c
1
C -N=C
I
I
O=
H R '
I
c
205
I
210
der favorable conditions, 34 may be formed in yields of 5 YO
based on the consumed pentose, it becomes clear that this
compound plays a central role in color formation. The pyranone 206 is a further colored product derived from pentoses.[1341Both 3- and I-deoxydicarbonyl compounds can be
formulated as precursors of 206.
I
+
H,N-CH-CNH-
cI = o
I
c=o
I
k
I
*
1
C-N=C-CNH11
II
C-OH
0
211a
-
I
I
1
C = N -C =CNHI
OH
C=O
I
211b
3.4. Shelf Life
Cleavage products not only participate in the formation of
volatile Maillard products, but are also found in the structures of colored compounds. The yellow pyrroles 207 and
208 were isolated from reaction mixtures of primary amines
with dihydroxyacetone, pentoses, and h e x o s e ~As
. ~expect~~~
ed, the concentration of such pyrroles increases in these sugar reaction mixtures when the pH is raised above 7 . Dihydroxyacetone and methylglyoxal, formed on loss of water,
occur in smoke, so that browning products, such as 207 and
582
It has been known for some time, that beer is stabilized
against oxidative changes through substances formed by
Maillard reactions occurring during the kilning of germinated barley. Without at first knowing their structure, these
compounds were called reductones. The triosereductone 212
and reductic acid 213 can be prepared from sugars, but the
conditions required (heating in alkaline or acid solution)
have no relevance to foodstuffs. It is now known that it is
reductone ethers of type 34, 48,51, 59, and 62 and aminoreductones 68, 69, 90, and 101, structurally comparable with
vitamin C, that act as stabilizers. The formation of reductones in the Maillard reaction has already been dealt with (cf.
Angew. Chem. I n f . Ed. Engl. 29 (1990) 565-594
Section 2.5.2). The compounds which are preferentially reduced by reductones are only just being identified. If ascorbic acid or acetylformoin 62 reacts with diacetyl, some 3-hydroxybutan-2-one is formed, i.e., diacetyl is reduced.['401By
analogy. 1-deoxydiketoses should lead to the reduced substances 51, 214,[611215,['411216,['401217,['421and 218,"431
as well as to the oxidized substances 219['411and 220.['441
storage of the pyranone 59 leads to larger amounts of the
oxidation product 220,['44] a compound that has also been
detected in heated or stored foodstuffs. The mechanism of
this oxidation has not yet been ascertained, but the disproportionation of a dimeric intermediate does not seem unreasonable. Oxidation of 59 with iodine or cupric salts gives
predominantly glyceric acid and lactic
$4
OH
$.=o
--
HO-C-
6
+
6a
C-OH
1
c=o
c=o
HC-OH
HC-OH
I
I
I
y
-
C-OH
C=O
I1
C-OH
C=O
I
+
I
HC-OH
C=O
I
HC-OH
HC-OH
H,C- OH
H,C- OH
I
FH3
I
I
c=o
c=o
1
I
7%
HC - OH
H,C-OH
21L
215
219
216
In comparative tests for antioxidant activity, aminohexosereductones 68a perform particularly well, though their relatively high toxicity does not commend them for use in foodstuff~.['~
For
~ ' detection of reductones in reaction mixtures,
two methods are available: selective electrochemical determination after separation by column chromatography['481
or condensation with hydrazine to hydroxypyrazoles of type
225, which can be determined gas chromatographically after
derivatizati~n.['~~I
Carbocyclic reductones cannot be con-
I
cI = o
217
51
218
220
C-OH
I1
NH,-NHCH,
-+
c-x
I
Following the same principle, one would expect the
pyrrolinone 218 to be formed directly from the aminoreductone 69a. Loss of water and addition of a second molecule
allows disproportionation to 218 to take place. This mechanism at least affords an explanation for the formation of 218
upon storage of 69a.1143]
The aminohexosereductones 69 a in acid solution give
good yields of the so-called methylenereductinic acid
222.['451Evidence for 222 in sugar-amine reaction mixtures
and in foodstuffs has not yet been obtained; however, rapid
condensation with carbonyl compounds to give cyclopentenediones of type 223, as well as ready dimerization to
221['461provide grounds for thinking that the methylenereductic acid is too reactive to be readily detected in reaction
mixtures.
221
-cy
222
223
On warming the hydroxyfuranone 51, some dimer 224 is
formed. presumably via radical intermediate^.['^] Longer
AnRew Chem. 102 11990) 565-594
Q
74
3
X = OH,OR,N,
,
HoG&H,
225a
+
H(3qrH3
225b
verted into hydroxypyrazoles. On treatment with methanol/
toluenesulfonic acid, volatile methoxy derivatives are obtained.1'411 Although it has been shown that certain
melanoidins possess antioxidative properties,[' 501 there is no
experimental evidence for the assumption that reductone
groups are responsible.
3.5. Mutagenic Substances
Since the introduction of the Ames test for mutagenicity,
a series of reaction mixtures, fractions, and Maillard products has been subjected to such a test. The results of experiments with foodstuff^,^'^ 'I aroma concentrates,[' 521 model
systems," 531 and fractions derived from them are usually not
comparable and are in part contradictory. At present, attention is focused on the Maillard products 172 and 173, the
carcinogenicity of which has meanwhile been thoroughly
substantiated." 541 Rapid and definite detection of these
compounds, which occur in some foodstuffs in ppb and less,
still presents certain problems. It has been indicated already
in Section 2.8.5, that steps have been taken to greatly reduce
their formation. The mutagenic and carcinogenic tryptophan derivatives 150 and 151 also have a certain significance for
Mixtures of these compounds with
melanoidins isolated from reaction mixtures of glucose and
glycine definitely exhibit reduced rates of mutation in the
Ames
The practical implications of such results cannot yet be assessed.
583
The mutagenic and in part strongly carcinogenic effect of
nitrosamines has been known for some time. Secondary
amines present in foodstuffs can react with added nitrite
during manufacture or preparation and with nitrite from
saliva as they pass through the stomach. Secondary amines,
such as Amadori compounds, are formed during the Maillard reaction. After addition of nitrite, except for the tryptophan derivative," 56a1 no mutagenic activity is found with
Amadori and Heyns compounds," 56b1 but the tetrahydro-pcarbolines of postulated structure 226r'571and nitrosothiazolidines of structure 227 exhibit unequivocal mutagenicity." "I However, the conditions of reaction, including the
nN N O
S
226 R = H, CH20H
r
227
concentrations of the reagents, need to be considered. For
example, the incubation of soy sauce with an amount of
nitrite corresponding to the maximum amount to be expected in saliva, under conditions equivalent to those in the
stomach, does not lead to an enhanced rate of mutation in
the Ames test, even though compounds 142 and 143 are
present in this food in relatively high amounts (50 and
10 ppm, respectively)." 591
Recently, the anticarcinogenic activity of Maillard prodIt is postulated that reactions
ucts has been
take place with oxygen radicals, which are trapped and thus
converted into less aggressive compounds.
As well as other compounds, the Maillard products 18,
25a, and methylglyoxal have been shown to be mutagenic in
the Ames test, while 145 and 146 have been shown to be
comutagenic.[16'] If, for example, one wishes to avoid the
formation of compound 18 in foods, one would have to be
prepared to substantially forego cooking.
3.6. Reduction in Nutritive Value
The growth of animals, which have been fed with proteins
that have been heated in the presence of sugars, may be
retarded. Many such studies have been carried out with milk
proteins. From these, it became apparent that reaction of the
sugar with the E-amino groups of the Iysine side-chains partly leads to products which the animal can no longer utilize.
This is equivalent to reducing the nutritive value, since lysine
is an essential amino acid. It has been shown unequivocally
that the formation of the Amadori compound 3c leads, after
proteolysis of the protein in the alimentary canal, to the
lysine derivative 3a, which is worthless to the organism.['621
One of the consequences of these results has been that, in the
production of milk powders and other products preferred in
infant nutrition, particular attention must be paid to keeping
damage to lysine to the minimum.
Amadori compounds possess good complexing properties, which lead, for example, to increased zinc excretion via
urine after parenteral"
or
administration of such
compounds. Experiments with oral intake of reaction mix584
tures showed that products subsequent to the aminoketoses
exhibit similar effects (cf. Substance 71).[1641
The resorption rates of higher molecular weight Maillard
products from reaction mixtures of pentoses and hexoses
with amino acids is very low in rats;['65]browning mixtures
from proteins and sugar are somewhat better absorbed, the
enzymatic hydrolysis of protein in the alimentary canal being
hindered by the Maillard reaction, but not blocked.[' 661
3.7. Control of the Maillard Reaction
When sterile solutions of glucose and amino acids are
mixed in the cold in the preparation of infusions for parenteral nutrition and stored with refrigeration at 2 "C, evidence of the formation of aminoketoses of type 3 can be
obtained in two to three weeks, and, after two months, a light
yellow coloration becomes apparent." 'I A parallel situation
occurs in many foodstuffs. If the Maillard reaction is to be
suppressed, this can be achieved, for example, by adding
~ u l f i t e s ~ or,
' ~ ' under
~
certain circumstances, by reducing the
water content/activity.['681However, such methods are not
appropriate for all foodstuffs.
It has long been known that sulfurous acid hinders the
Maillard reaction. One can assume that addition of sulfite to
the carbonyl group occurs,[169 which thus becomes
blocked and is no longer available for further reactions.
These reactions are not always reversible, since part of the
sulfite remains bound, for example, in the higher molecular
weight reaction products.[169b1 One of the few products isolated and identified to date has the structure 228.[169a1
It
HC=O
c=o
FH2
HC - S03H
228
HC-OH
HC
,
- OH
seems reasonable to assume that the intermediate 17 undergoes reaction with sulfite. The Maillard reaction is also hindered by the addition of divalent sulfur derivatives, such as
thioglycollic acid or cy~teine.["~~
Since addition of sulfurcontaining compounds is expected to lead to large alterations in aroma in foods, their practical application is hardly
likely.
Control of the Maillard reaction in vivo by means of
aminoguanidine will be discussed in Section 6.2.
4. Formation of Amadori Compounds
under Physiological Conditions
4.1. Reactions of Glucose with Proteins
That the Maillard reaction also occurs in vivo only became clear when the heterogeneity of hemoglobin was discovered. Thus, it was found[1711 that, when human
hemoglobin is chromatographed on weakly acid ion exA n g e w Chem. Int. Ed. Engl. 29 (1990) 565-594
changers, three hemoglobin fractions (HbA,,, HbA,,, and
HbA,,) elute in front of the main fraction, HbA, (see Fig. I).
These fast hemoglobins (designated HbA, as a group),
which normally constitute a few per cent (5-8 %) of the total
hemoglobin, possess the same amino acid sequence as HbA,.
They differ only through the presence of additional, cova-
represented in a general way in Equation 1, glucose (G)
reacts with the free amino groups of a protein (P), forming
a Schiffs base 1. After rearrangement to the relatively stable
aminofructose 312, the fructosylated protein persists in the
body until it is degraded.[‘761The amount of fructosylated
protein present in the body at a particular time represents the
steady-state concentration [3clSs.
G
k
+ Pd
1
k
I
HbA 1
3 c 5 degradation products
Under steady-state conditions, when k - , 4 k ,
tion (2)
(1)
Equa-
HbA,
“bA 1c
A LO5
k,
k-2
M
The rate constant for the forward reaction is given by Equation (3),
Fractions
Flg. 1. Elution profile of human hemoglobin after chromatography on a weakly acid ion exchanger. Hemoglobin fraction HbA,,, in which glucose is bound
as structure 3b, elutes before the main fraction HbA,. Recently, it was shown
that fraction HbAo also contains fructosylated hemoglobin, in which, however,
glucose is bound as structure 3a [172a]. The composition of the hemoglobin
fractions HbA,, and HbA,, remains in part unknown. The shaded portions
indicate increases with increased blood glucose concentration (diabetics).
A,,, = absorption at 405 nm.
lently bound, low molecular weight substances. The elucidation of the structure of HbA,, aroused particular interest, as
higher amounts of it were found by chance in the blood of
diabetics.[’ 731 With the aid of periodate oxidation, it could
be shown that the hexose found in HbA,, is bound at the
N-terminal valine of the P-chain in the form of l-amino-ldeoxyfructose 3b. Since HbA,, can also be obtained by in
vitro incubation of HbA, with glucose,[’751 it is clear that the
formation of HbA,, proceeds via the addition of glucose to
the N-terminal amino group of hemoglobin, forming the
Schiffs base 1, and subsequent Amadori rearrangement to 3.
The reaction has been termed the non-enzymatic glycosylation of proteins, since, in contrast to the multitude of enzymatically formed glycoproteins, glucose is here bound to
protein without mediation by enzymes. In the American literature the synonymous term ‘glycation’ also appears.
In medical circles, ‘fructosamine’ is used as a synonym for
3. Here, since the Amadori compound 3 formed is always a
l-amino-2-deoxyfructose, the term ‘fructosylation’ will always be employed. With hemoglobin, the preferred point of
fructosylation is the amino terminal of the p-chain. The amino terminal of the a-chain is fructosylated 8 - I0 times less
frequently in c o m p a r i s ~ n . ~Of’ ~the
~ ~ 22 primary amino
groups of the a- and p-chains, only 3 are preferentially fructosylated. It is noteworthy, that the non-enzymatic fructosylation shows such an unexpected specificity, in which the pK,
value of the amino groups apparently plays a subordinate
role. Limited steric access for glucose to the less reactive
amino groups has been discussed as a possible reason for the
specificity, but, meanwhile, indications have been obtained
that the specificity could also be explained through acid-base
catalysis by adjacent proton donor/acceptor groups.[’721As
Angew Chem. 102 11990) 565-594
(3)
and that for degradation is k , . The protein half-life can be
calculated by means of the expression, t l i z = ln2/k,. The
concentration of fructosylated protein (degree of fructosylation) in the steady state is therefore proportional to the glucose concentration and the ratio of the formation and decomposition constants. Since the degree of fructosylation in
the steady state and the rate constant k; can be determined
for each protein, one can calculate the decomposition constant and hence the protein half-life, provided the glucose
concentration is known and constant (about 4.4 mM in the
human body). On the other hand, if the rate of decomposition is known, the degree of fructosylation of a protein becomes a measure of the glucose concentration to which it has
been exposed during its ‘life’. In addition, the following conditions must also hold: 1 . The protein must be freely accessible to the glucose in vivo (which is so for the blood proteins
and the proteins of insulin-insensitive cells, where inward
transport of glucose is greater than its consumption). 2. The
in vivo and in vitro kinetics must correspond. Fructosylation
can thus be regarded as an endogenous, non-radioactive way
of marking proteins, which can be used for the estimation of
the half-life of proteins.
If one determines the steady-state degree of fructosylation
of human proteins that have been separated from the serum
of normal subjects, and measures the corresponding in vitro
rate constants k;, the protein half-lives can be calculated.[1761Some results are presented in Table 1. The values
given in the literature[’771show good agreement with the
values obtained by in vitro kinetics for serum proteins. The
low value obtained by the latter method for hemoglobin is
associated with the fact that hemoglobin is present in red
blood corpuscles not in solution, but as a concentrate
(330 mg/mL).
Under conditions of increased or decreased protein
turnover, as, for example, with thyroid hyper- or hypo-activity, respectively, a decreased or increased degree of fructosylation was indeed found for the human serum lipoprotein,
‘low-density lipoprotein’ (LDL), which is engaged in choles585
Table 1. Comparison of the half-lives tlil of proteins with data from the literature. The values for the rate constant k', and the degree of fructosylation in the
steady state were determined in vitro.
Mean degree
of fructosylation
nmol lys-fru mol lys-fru
[a] per mg
per mol
protein
protein
Albumin
y-Globulin
Fibrin peptide
LDL [c]
HDL [dl
Hemoglobin
3.6
1.6
0.86
1.2
1.8
2.0
k;
[mM"
d - ' mg-'1
Calcd.
[b]
0.04
0.017
0.048
0.04
0.05
0.023
14.04
14.6
3.0
4.68
5.62
13.6
0.25
0.25
0.2
0.5
0.1
0.13
tration of normal subjects (4.4 mM), i.e., the in vitro reaction
conditions appear to reflect those in the body relatively well.
11:2
Ref.
11771
12-20
15-26
4-5.5
3.1 -3.4
4.2-4.8
70
[a] Lysines reacted at c-amino groups; structure 3c. [b] From f, = InZ/k,
[c] Low density lipoprotein (cholesterol-transporting protein). [d] High density
protein (cholesterol-removing protein).
4.2. Comparison of the Structure of Enzymatically Glycosylated and Non-enzymatically Fructosylated Proteins
In many glycoproteins, different sugars (e.g., galactose)
are bound 0-glycosidically to serine or threonine present in
the peptide chain (cf. structure 229). For a long time, only
one type of N-glycoside had been characterized, namely,
N-acetylglucosamine 781bound N-glycosidically to the amide nitrogen of asparagine (cf. structure 230).
0
terol transport.['76b1 Particularly interesting was the observation that the degree of fructosylation of LDL of normal
subjects was correlated with the LDL-cholesterol level : a
doubling of the level is reflected in a doubling of the degree
of fructosylation of the LDL1176
b1 (cf. Section 5.2). In Figure
2, an experiment is presented in which a model reaction
simulates the in vivo protein turnover.[176b1 An albumin so-
B
10
20
30
t [dl
-
NHCoCu,
229
4="
230
&Q
$
In contrast, the fructosylation of proteins takes place
mainly on the free amino groups of the lysine residuesr179]
(cf. structure 3c and Section 4.1). The conformation of fructosylated proteins in aqueous solution has been examined by
I3C-NMR spectroscopy.[1801RNase A from bovine pancreas was used as the model protein. The results showed that
the protein-bound hexoses are almost exclusively present in
cyclic form, as would be expected by analogy with the free
hexoses. The glucosylamine 1, as well as the ketosamine,
prefers the /3-pyranose form (see structural formulas in Section 2.4). Since the reference compounds, N-a-formyl-E-fructoselysine and fructosylated poly-L-lysine, exhibit resonances corresponding to those of the fructosylated enzyme
(RNase), it seems that the protein matrix has relatively little
influence on the conformation of the hexose.
LO
Fig. 2. Simulation of human protein-turnover in vitro. Human serum albumin
incubated with 7.2 (A) and 14.4 (B) mM o-glucose. respectively. Each day, a
twentieth part of the total was removed and replaced by the same amount of
further original medium. Additional details appear in the text.
lution was incubated with two different concentrations of
glucose and daily a twentieth of the solution was replaced by
starting solution. This daily 'elimination' of 5 % corresponds
to a half-life of 13.8 days in a first-order degradation reaction. The border concentration between normal and raised
blood-sugar level was taken to be 7.2 mM and the concentration for diabetics 14.4 mM glucose. After 40 days, equilibrium is more or less achieved. The steady-state value for
7.2 mM glucose was 5.1 nmol Iys-fru per mg albumin, whereas, with the doubled glucose concentration, roughly double
the degree of fructosylation was obtained (lys-fru indicates
structure 3c). From these values one can estimate that the in
vitro steady-state fructosylation of 3.6 nmol albumin required a glucose concentration of about 5 mM. This concentration of glucose agrees well with the mean glucose concen-
586
N
Iu
NH-C-CH,-CH
4.3. Methods for the Quantitative Determination of Nonenzymatically Fructosylated Proteins
Since many of the methods used in food chemistry to
detect and determine in vivo fructosylated proteins are not
specific and sensitive enough, a series of alternative methods
has been developed. Most of the methods of detection depicted in Figure 3 do not proceed stoichiometrically ; they
constitute therefore not absolute, but only relative methods
of determination. Theoretically, it should be possible to
quantify 3, bound in a protein, after hydrolysis of the
protein. However, 3 is not stable under the strongly acid
conditions of hydrolysis; one obtains about 50% free lysine
and about 30% 82a (cf. Section 2.5.4). It could be shown[641
that 82a, determined by means of an amino acid analyzer,
can be used to measure the degree of heat damage suffered
by foodstuffs, especially milk products, through non-enzymatic browning. Considerably smaller amounts of 82a can
be determined by HPLC.1181JUse of this method allowed
50 pmol3c in 5-25 pg protein to be determined specifically
and precisely. A further advantage of the method is that it
Angew. Chem. Int. Ed. Engl. 29 (1990) 565-594
82a
t
Antigenantibody 4
\
6 N HCI. tOO',C
Antibody
/
Electrophoresis,
chromatography
2 -
/
18
OH e
\
Oxidation p r o d u c t s
Fig. 3. Reactions providing evidence of the presence of fructosylated proteins.
The reactions on the right occur with modification of structure 3, those on the
left without such modification.
permits insoluble tissues (e.g., collagenous ones) to be investigated.
Fructosylated proteins can also be determined in the amino acid analyzer after sodium borohydride reduction to 231
and subsequent hydrolysis, in which 231 dehydrates to a
NH - CH-CO
I
(FH,)'
NH
I
other reducing substances, present in human serum, have
substantially completed their reactions within the first ten
minutes. The alteration in the pK, of hemoglobin fructosylated at valine 3b, which had led to the discovery of the
fructosylated hemoglobins, was used at first by all researchers to separate and quantify 3b by means of ionexchange chromatography, electrophoresis, or isoelectric
focusing.'' "1 Since, as already mentioned, most proteins are
mainly fructosylated at the &-aminogroups of lysine residues
(cf. structure 3a), which exhibit substantially less change in
pK, these methods are not universally applicable (cf. Fig. 1).
Attempts to detect 3 with phenylhydrazine as a carbonyl
reagent failed, because of insufficient reactivity and sensitivity,[1861
Although some protein-bound sugar structures exhibit extraordinary antigenicity, early trials showed already that
structure 3 becomes antigenic only after reduction to 231.
However, the binding of the antibody to 231 is not quantitative for all fructosylated proteins, probably because of steric
hindrance, so that the method is not universally applicable.[1871
Fructosylated amino acids and peptides bind specifically
to matrix-bound boric acid, with formation of a boric ester
(see Fig. 3), and can be quantified after elution with acid or
sorbitol.1'881However, this procedure also is not quantitative for many proteins and therefore not appropriate, probably again because of steric hindrance. Our own investigations have shown that boric ester formation can also take
place with substituted phenylboric acids in homogeneous
aqueous solution.[189]Since the absorption maximum of 232
moves to shorter wavelengths on esterification, 3 can be
quantified photometrically.
?HZ
HC-OH
HO-CH
HC-OH
HC-OH
I
HF-OH
231
considerable extent. Usually the sensitivity needs to be improved, for which purpose tritiated sodium borohydride is
employed. The determination of 18, which is frequently used
in food chemistry because it is readily formed from 3 on mild
acid hydrolysis (cf. Section 2.5.2), can only be applied here
under special circumstances.[182.1831
The characteristic behavior of Amadori compounds on
reduction was employed for their detection relatively early
6l In alkaline solution, 3 enolizes to the strongly reducing enaminol5, which can be determined and quantified by
means of various redox indicators, such as dichlorophenolindophenol. Since the reductone formed also possesses reducing properties, the reaction without, firstly, unequivocal
stoichiometry and, secondly, a clearly defined endpoint.
To overcome these problems, it was suggested that the
reaction be determined kinetically with Nitrotetrazolium
Blue as the reducing agent.['841 It became evident that at
pH 10.3 neither glucosylamine nor free glucose participated
in the reaction to any marked extent. By appropriate choice
of time window, the specificity could be improved, because
Angru.. Clwm 102 11990) 555-594
It can be seen that, in spite of the number of methods
described, there is no generally applicable technique for determining precisely and specifically proteins that have been
fructosylated in the human body. For each particular
protein, the most appropriate method has to be selected from
those available.
4.4. Detection of Fructosylated Proteins in Human
Blood and Tissues
Only in this decade was it realized that many human
proteins, not just hemoglobins, are fructosylated. About ten
years ago, it was shown['901that human serum albumin is
also fructosylated. As with hemoglobin, the position of fructosylation is relatively specific. Only nine of 55 &-amino
groups are measurably fructosylated, LYS-525 being most
strongly attacked (50% of the total fructosylation). The degree of fructosylation of some blood proteins is listed in
Table 1. Since diabetics have higher blood-sugar values
(about 2-5 fold), some proteins from diabetics may bind
more than 1 mol fructose per mol protein. Other plasma
587
proteins, such as a,-antitrypsin, fibronectin, plasminogen,
antithrombin-111, ApoE, and ApoC 11, can of course be determined, but, mainly because of their short half-life (hours,
1-3 days), they are hardly f r u ~ t o s y l a t e d . [ ' ~Not
' ~ only soluble proteins, but also membrane-bound proteins are subject
to fructosylation, as has been demonstrated with erythroc y t e ~ . [ ' With
~ ~ ] human tissue proteins, the degree of fructosylation depends, amongst other factors, on whether glucose
has ready access to the proteins (as with extracellular
proteins such as connective tissue) o r whether (as with intracellular proteins) the glucose concentration in the cell is very
low. This is particularly relevant for insulin-sensitive tissue,
i.e., tissue in which glucose uptake is regulated by insulin. As
a consequence, fructosylation could not be detected in the
proteins of the muscles of the heart o r ~ k e l e t o n . [ ' ~ ~ 1
In contrast, in insulin-insensitive tissue, such as nervetissue, 3c could be detected.['931 Tissue containing collagen
has been particularly thoroughly investigated. Figure 4
shows values for the content of 3c in tendon, aorta-wall,
specific methods, showed a smaller increase (about 50 Yo)in
the fructosylation of tendon and aorta tissue." b1 However,
a very clear relationship between degree of fructosylation
and age has been established for lens tissue of the
Since, according to present knowledge, fructosylated amino acids cannot be metabolized in the mammalian organism
(cf. Section 3.6) and since they have good solubility in water,
they should be eliminated largely in the urine. In fact, normal
subjects d o excrete significant amounts of fructosylated amino acids, about 0.1 -0.3 mmol per day.['88,'961As some
foodstuffs contain substantial amounts of free o r bound
fructoselysine, 3a and 312, respectively, the question arises
whether the excreted fructosylated amino acids are not in
part of exogenous origin. Evidence that this IS not so comes
from the fact that fasting for 24 hours leaves the excretion
rate unchanged, fasting for I0 days even increasing it.
This increase is due to heightened protein degradation
(catabolism). With animal experiments it has been shown
that exogenous fructosylated amino acids and proteins are
degraded by microorganisms in the gut. Mammals cannot
make use of fructosylated amino acids. Diabetics excrete
increased amounts of fructoselysine in their urine, corresponding to their metabolic state.[188.1961
''
z
4.5. Do Subsequent Reactions also Occur in vivo?
Formation of 'Advanced Glucosylation End Products'
(AGE)
I
..
T
3
f
2%
I
L
&u=
0
0
1 +.--
I
I241
1\21
Tendon
1
I211
,131
!
Aorta
[Media intirnol
1101
('1
Coronary
arteries
'151
'8)
Peripheral
nerves
It1
'5!
Lung
connective
tissue
:
181
Glomerular
basement
membrane
Fig. 4. Non-enzymatic fructosylation of several ~ I S S U ~from
S
diabetic ( 0 ) and
normal ( 0 )patients. The horizontal lines indicate the arithmetic means. The
number of cases is given in brackets for each type of tissue. Fructosyidtion is
expressed in nmol lys-fru per pmol phenylalanine, which serves as an arbitrary
measure of the amount of protein. The glomerular basement membrane is the
collagenous dialysis membrane of the kidney.
nerve tissue, lung connective tissue, and glomerular basement membrane. 3c has also been detected in hair and in
finger nails (cf. Section 6.1). The mean values for the different tissues of normal subjects correspond roughly to a fructosylation of 1-3 YO of all the lysine residues. The mean
values for individual diabetics are more variable than those
for non-diabetics, corresponding to greater differences in
metabolism. The degrees of fructosylation of the different
tissues correlate well with each other: patients with high
fructosylation of tendons also exhibit high fructosylation in
other tissues. Mean blood sugar levels also correlate surprisingly well with the degree of fructosylation of tissues.
Of particular interest is the question whether the degree of
fructosylation of proteins, especially that of connective tissue, increases during life. Although one paper" 941describes
a doubling of the degree of fructosylation of human tendon
collagen during a lifetime, other researches, using more
588
181
Whether the Amadori product 3, in a way corresponding
to the browning reaction of food chemistry, also reacts further in vivo and so provides a possible mechanism of aging
is an interesting problem. The first support came from finding that the fluorescence specific for browning pigments increased for tendon collagen in a linear manner with age.['971
Samples obtained from diabetics did not give readings falling
into the normal range, but exhibited greater fluorescence.
Since 3 does not fluoresce, derivatives must have been
formed. In model reaction mixtures of bovine serum albumin
and glucose, a corresponding fluorescence was clearly present after 30 days.['981From the solution obtained from the
protein after acid hydrolysis, the fluorescent product 88 was
isolated. As already outlined in Section 2.5.4, it is probably
an artefact. Meanwhile, the pyrrolealdehyde 25b has been
shown to be present in a number of native proteins by means
of a specific antibody, thus proving the presence of derivatives of 3 in V ~ V O . [ ~The
~ ] mechanism of formation has already been described in Section 2.5.1.
A further derivative, carboxymethyllysine 128, also
formed in vivo, has been detected in the meantime (cf. Section 2.6.2). This therefore explains the excretion of 128 in
urine, which had been known for some time.['991128 has also
been found in the eye lens, where its concentration increases
linearly with age." 991 The interesting observation has been
made that in vitro 3a leads to more browning pigments in the
absence of oxygen than in its presence, but without formation of 128.[771
It was therefore proposed that the excretion
of 128 can be used as a general indicator of the age of extracellular protein and the ratio of excreted 3a/128 as a general
indicator of oxidative stress in the human body."991 Meanwhile, in a model reaction, a further pyrrolealdehyde,[2001
39,
Angcw. Chem. Inr. Ed. Engl. 29 (1990) 565-594
has been found, the mechanism of formation of which has
already been postulated (cf. Section 2.5.1). In the presence of
sodium sulfite, compound 39 accumulates, while, in the absence of sulfite, it very quickly forms cross-links with the free
amino groups of proteins. One can assume that compounds
of type 233 are formed.
Meanwhile, the Maillard product 167 has also been isolated from human
The structure of this cross-link
implies participation of lysine, arginine, and a pentose.
Since, so far, nothing is known of the role of arginine and of
pentoses in the Maillard reaction in vivo, the significance of
167 for the cross-linking of proteins cannot yet be evaluated.
Recently, it was demonstrated that the Maillard reaction
is involved not only with proteins, but also with nucleic
acids.[2011Incubation of DNA with reducing sugars leads to
the formation of brown fluorescent substances, the spectroscopic properties of which resemble those of AGE-proteins.
Such modified DNA (AGE-DNA) on transcription exhibits
specific mutations, caused by insertions or deletions. Crosslinking between proteins and DNA was also evident.
Whether such protein-AGE-DNA cross-linking also takes
place in human cells has not yet been demonstrated, but, on
the basis of the above findings, seems not unlikely.
molecules. The attack of glucose on a lysyl or hydroxylysyl
residue should inhibit the reaction and thus lead to reduced
cross-linking of collagen. In actual fact, the opposite was
found. To explain these experimental results, one must therefore draw on other mechanisms. As already described (cf.
Section 4.9, fructosylated lysyl groups can themselves form
cross-links via derivatives such as 39 and 167. Such nonphysiological cross-links are formed in collagen fibers between collagen molecules lying alongside one another, thus
increasing the mechanical stability, but lowering the solubilization by acids or enzymes. In model experiments, it has in
fact been shown that tendon has greater mechanical strength
after incubation with glucose. Incubating tendon with glucose simulates the increase in tensile strength which takes
place with age.[2o41Using RNase as an example, it has been
shown that this enzyme, on fructosylation in vitro, undergoes di- and trimerization in a time-dependent manner.[2o51
This cross-Iinking continues even after removal of the glucose and can be almost completely inhibited by simultaneous
incubation with L-lysine. The reaction mechanism was considered to involve condensation of an amino group of the
protein with 3a of another protein, thus producing crosslinks (structure 13). Meanwhile, it has been shown that such
cross-linking can also take place between different
proteins.['''. 'O61 When serum albumin or immunoglobulin
G is added to collagen, previously incubated for a longer
period with glucose, they become covalently bound. In fact,
with diabetic animals, five times more immunoglobulin G is
bound to collagen than with non-diabetic ones. Particularly
interesting is the discovery that the cholesterol-transporting
serum lipoprotein LDL can be bound covalently in this manner to fructosylated collagen (e.g., in the arterial wall).
5.2. Effects of Fructosylation on the Function of Proteins
5. Does the Maillard Reaction Lead to Modification
of the Structure and Function of Proteins?
5.1. Effects on the Structure of Proteins
Fructosylation of bovine lens protein leads to a dulling of
crystallin.[2021Since this opaqueness does not appear in the
presence of reducing agents, it is assumed that fructosylation
gives rise to a subsequent change in conformation, exposing
sulfhydryl groups and thus allowing dimerization of the
protein. In cataracts (lens dulling) from diabetic and galactosemic rats, evidence was obtained for aggregates of crystallin. linked by disulfide
As, according to other
studies, crystallin is fructosylated in a time- and concentration-dependent manner, one is led to assume that the Maillard reaction piays a role in the development of cataracts
with age or diabetes.
Since cross-linking of the free lysine sidechains of the longpersisting connective tissue proteins, collagen and elastin, is
of essential significance, the relationship between the fructosylation of collagen and its properties has been extensively
investigated. During physiological cross-linking of collagen,
a lysyl oxidase oxidizes a few lysyl and hydroxylysyl residues
to the corresponding aldehydes, which subsequently crosslink with the &-amino groups of neighboring collagen
Fructosylation occurs primarily on the free &-amino
groups of lysine bound in proteins. These amino groups of
lysine are of essential significance in the active centers of
numerous proteins with functions in catalysis, transport,
binding, and structure. They also play an important role in
interactions, such as those between hormone and hormone
receptor, antigen and antibody, and enzyme and enzyme
inhibitor. Hence, it is an attractive hypothesis to postulate
that fructosylation of lysyl residues, occupying strategically
important positions, is detrimental to the functionality of the
protein affected. However, when transaminases were fructosylated in vitro, no reduction in enzyme activity was
found.['891Evidently, the lysine at the active center, which
forms a Schiff s base with the cofactor, pyridoxal phosphate,
must be protected from fructosylation. In contrast, on incubation with glucose, ribonuclease A loses 50 % of its original
activity in 24 hours, whereas cathepsin B and papain lose
about 70% in 2 weeks.[2071With these enzymes, it appears
that fructosylation of the lysine, which lies close to the active
center, is responsible for the reduction in activity.
Changes in conformation are also possible factors, as has
been shown with the ~-N-acetyl-D-ghcosaminidase from
bovine kidney as example.r208JThe activity was already reduced to a half after seven days' incubation with 4 4 m ~
D-glucose. Studies with the separated isoenzymes A and B
589
demonstrated that isoenzyme A is completely inactivated,
whereas B remains fully active. The changes in functionality
are accompanied by alterations in structure. Thus fructosylation changes the electrophoretic mobility. The molecular
mass increases from 130 to 205 kDa. To what extent these
reductions in activity and changes in structure play a role
also in vivo is so far not known. A series of experiments
indicates that fructosylated proteins are less readily split on
proteolysis. Thus, it appears that the cross-linking of protein
chains, which occurs during the course of the non-enzymatic
browning reaction between reducing sugars and proteins,
also affects their ease of hydrolysis (cf. Sections 3.6 and 5.1).
The proteolytic cascade involved in blood clotting has been
the subject of research by many groups. The inhibitory effect
of antithrombin 111 is considerably reduced on incubation
for 3 days with 100 mM glucose; 400 mM glucose reduces it to
half. Such reduced functionality of antithrombin I11 leads to
higher thrombin activity and to increased formation of fibrin
peptide. The greater the degree of fructosylation of the fibrin
peptide, the more slowly is it cleaved by plasmin.[2091Hormone activity can also be affected by fructosylation, as can
be seen, for example, with insulin. The diminished activity of
insulin can be demonstrated on fat cells through reduced
glucose oxidation, increased de novo fatty acid synthesis,
and reduced antilipolytic properties.['"] Since these effects
only appear with very high degrees of fructosylation, it is
likely that these results are not relevant to the situation in
vivo. The cholesterol-transporting L D L of human blood
normally contains about 0.5 mol lys-fru per r n ~ I . [ ' b1
~ ' Detailed research in vitro showed that modification of the 6amino groups of LDL led to a reduction in binding to the
LDL-receptors, which are attached to the cell membrane.'" 'I The reduction in binding, uptake, and
metabolism of LDL by different human cells could be
demonstrated, also for fructosylated LDL. to be dependent
on the amount of glucose bound. It is noteworthy that fructosylated LDL in human macrophages can stimulate cholesterol ester synthesis. It seems possible, therefore, that fructosylation of L D L and the corresponding derivatives (AGE
products) favor the formation of foam cells (macrophages
are taken to be precursors of the foam cells of the atherosclerotic plaques). The relevance, in this connection, of the interesting observation that raised LDL-cholesterol levels are accompanied by increased degrees of fructosylation (cf.
Section 4.1), is so far not known.
5.3. Immunogenicity of Fructosylated Proteins
and their Reaction Products
Several attempts to detect unequivocally antibodies
against structure 3 in human blood have thus far not succeeded. This corresponds to the results of animal experiments, which indicate that fructosylated proteins possess little antigenicity (cf. Section 4.3). Since the amount of A G E
products, which has been found in the different long-life
body proteins, is much less than had been expected on the
basis of in vitro studies, the possibility was investigated of
the existence of a mechanism in the living organism that is
capable of selectively removing AGE proteins. It was indeed
shown that, in immunocompetent cells (macrophages), there
590
is a high-affinity receptor, able to bind A G E proteins, to take
them into the cell, and to degrade them.[2121This receptor
for A G E proteins has already been extracted from the
macrophage membrane and concentrated. However, this
degradation system for AGE proteins does not seem to be
totally efficient, since the aging products accumulate over
the life span of the individual as a function of time and of the
blood glucose level.
6. Medical Aspects of the Maillard Reaction
6.1. The Maillard Reaction as Diagnostic of the
Metabolic State of Diabetics
The degree of fructosylation of proteins in the human
body, under steady state conditions, is proportional to the
half-life of the protein and to the glucose concentration (cf.
Section 4.1). Thus, provided the protein half-life remains
constant, the mean concentration of glucose to which that
protein has been exposed during its life time in the body can
be calculated. The degree of fructosylation can therefore be
used to calculate the blood glucose level retrospectively. The
fructosylated proteins represent a sort of blood sugar memory. Depending on the time frame one wishes to investigate, so
one selects a protein with an appropriate half-life. Thus, for
example, hemoglobin is suitable for checking on the blood
sugar level over the last 2- 3 months, whilst the determination of the fructosylated serum proteins mirrors the mean
blood glucose concentration over the last 2-3 weeks. By
means of the furosine method (cf. Section 4.3), raised glucose
levels can even be determined from the degree of fructosylation of finger nail or hair protein^.'"^^ The determination of
the fructosylated blood proteins thus leads to the objective
assessment of the state of metabolism, which is clinically
helpful both for the physician and the patient. The patient is
given a straight motivating indicator and the physician can
take into account the mean blood sugar values over a considerable time, without the effort o f having to determine the
individual values.
Since the late complications of diabetes appear, in particular, in the insulin-insensitive tissues, the fructosylation of the
proteins of these tissues was compared with the severity of
the late complications.['931Diabetics with severe late complications in several organs exhibited a high degree of fructosylation of tendon or aorta tissue. These results point to a
connection between poor metabolic control and the development of diabetic late complications. In accordance with this,
it could be shown recently that diabetics with high amounts
of AGE proteins suffer from increased damage to eyes and
kidneys.[' 14]
6.2. Intervention in the Maillard Reaction in vivo
Once the Maillard reaction had been recognized as the
cause of undesirable, glucose-dependent cross-linking of
proteins in the human body, ways of suppressing the reaction were sought. For many years, food chemists have employed sulfite as an additive to inhibit the formation of Maillard products in foods (cf. Section 3.7). For administration
Angen. Chem. Ini.
Ed. Engl. 29 (1990) 565-594
with 3 or other rearrangement products. It could be shown
that aminoguanidine 234 inhibits the formation of AGE substances and the cross-linking of proteins in vivo as well as in
vitro.[2 51 After administration of aminoguanidine, no accumulation was observed of immunoglobulins in the collagen
network of the blood vessels of diabetic rats. Further research showed lowered cross-linking of the collagen and less
deposition of the cholesterol-transporting lipoproteins on
the arterial walls. Studies of toxicity in different mammals
and in healthy humans provided evidence that 234 is relatively well tolerated. It is planned to test 234 as a drug to hinder
the cross-linking caused by the Maillard reaction and thus to
reduce late complications in diabetes and the damaging effects of aging.
It is assumed that aminoguanidine reacts preferentially
with 3 to form hydrazones of type 235. In in vitro experiments, the triazine 236 is formed to some extent. A probable
mechanism for its formation is the retroaldolization of 3 to
methylglyoxal 96, which then condenses with amino-
HN
,
+
3
- NH
H,N-C
I1
I
NH
234
H,C-NHNH
L
I/
C=N-N-C
I
H I
HQ-YH
NH2
HC -OH
HC -OH
235
HzC- OH
guanidine. Reaction of the hydroxyfuranone 62 with
aminoguanidine at pH 7-8 leads to formation of the hydroxytriazine 237.[216]
It can be assumed that 62 is cleaved to
give pyruvic acid, which then cyclizes with aminoguanidine
62
+
234
0
to give 237. The extent to which the formation of triazines
also occurs in vivo still needs to be ascertained.
7. Summary and Outlook
Even 78 years after its discovery, the Maillard reaction
with its multitude of reaction pathways and products is still
only known in outline. In food chemistry, the objectives are
to have the Maillard reaction proceed in such a way that the
formation of toxic substances and the reduction in nutritional value are suppressed, whilst, simultaneously, desirable
components are formed in optimal amount. Although it was
Angrii. c‘hcm. 102 (1990) 565 -594
first realized only relatively recently that the Maillard reaction also occurs in the human body, evidence has already
been provided for a series of structural and functional
changes in fructosylated proteins. However, in the foreseeable future, it will continue to prove difficult to show that,
from the confusing multiplicity of Maillard products,
specific, defined substances are responsible for the complex
processes of aging. Current efforts aim particularly to hinder
the undesirable Maillard reaction in the human organism
and thus the “Verzuckerung” of man and its consequences
during his or her life time.
We are greatly indebted to Prof. Dr. K . D. Gerbitz, Dr.
B. Olgemoller, and Dip1.-Chem. S. Prytulla for critically reading the manuscript and to Prof. Dr. Th. Severin for many
helpful comments.
Received: June 21, 1989.
revised: November 27, 1989 [A 762 IE]
German version: Angew. Chem. i02 (1990) 597
Translated by Professor H . E. Nursten. Reading (UK)
L. C. Maillard, C. R. Acad. Sci. Ser. 2, 154 (1912) 66; ibid. 155 (1912)
1554.
M. Amadori, Atti R. Acad. Nuz. Lincei Mem. Cl. Scr. Fic. Mat. Nut. 13
(1931) 72.
a) R. Kuhn, F. Weygand, Ber. Dtsch. Chem. Ges. B 70 (1937) 769; b)
F. Weygand, ibid. 73 (1940) 1259.
H. Simon, A. Kraus, Fortschr. Chem. Forsrh. 14 (1970) 389.
K. Heyns, H. Paulsen, R. Eichstedt, M. Rolle, Chem. Ber. 90(1957)2039.
H. Paulsen, K. W. Pflughaupt in W. Pigman, D. Horton (Eds.): The Carbohydrotes: Chemistry and Biochemistry, Vol. IB, Academic Press, New
York 1980, p. 881.
Ubersichtsartikel: a) T. M. Reynolds, Adv. FoodRes. 12(1963) 1; rhid. 14
(1965) 167; b) M. Angrick, D. Rewicki, Chem. Unserer Zcit i 4 (1980)
149; c) C. Eriksson (Ed.): The Maillard Reaction in Foods. Prog. Food
Nutr. Sci. 5 , Pergamon Press, Oxford 1981; d) G. Waller. M. S . Feather
(Eds.): Maillard Reactions in Food and Nutrition, ACS Svmp. Ser. 215.
American Chemical Society, Washington D.C. 1983: e) G Westphal.
L. Kroh, Nahrung 29 (1985) 757, 765; 0 W. Baltes, Leben.smittelchem.
Gerichrl. Chem. 40 (1986) 49; g) M. Fujimdki, M. Namiki. H. Kato
(Eds.): Amino-carbonyl Reactions in Food and Biological Swtcms. Dev.
Food Sri. 13, Elsevier, Amsterdam 1986; h) M. Namiki, Adv. Food Rrs
38 (1989) 115; i) J. W Baynes, V. M. Monnier (Eds.): The Maillard ReacZion in Aging, DiaberesandNutrition, Prog. Clin. 5101.
Re.% 304, A. R. Liss
Inc., New York 1989; j) T. H. Parliment, R. J. McGorrin, C.-T. Ho
(Eds.): Thermal Generation of Aromas. ACS Symp. Ser. 409. American
Chemical Society, Washington D.C. 1989, k) P. A. Finot. H U. Aeschbacher. R. F. Hurrel, R. Liardon (Eds.): The Maillard Reacfion in Food
Processing, Human Nutrition and Physiology, Birkhduser Verlag Basle.
Brussels, Berlin 1990.
a) R. Wittmann, K. Eichner, 2. Lebensm. Unters. Forsch. IRH(1989) 212;
b) H. Hashiba, Agric. B i d Chem. 42(1978) 1727; c) J. Steinig, A. Montag, Z. Lebensm. Unfers.Forsch. 174 (1982) 453; d ) K. Heyns. H. Paulsen,
Justus Liehigs Ann. Chem. 622 (1959) 160.
a) H. Sulser, W. Biichi, Lebensm. Wiss. Technol. 2 (1969) 105; b) M. Ciner-Doruk, K. Eichner, Z . Lrbensm. Unters. Forsrh. i68 (1979) 9, 360; c)
A. B. Moller, A. T. Andrews, G. C. Cheeseman. J. Dairy Rex 44 (1977)
267; d) H. Erbersdobler, Mikhwissrnsrhaft 25 (1970) 280.
H. Hashiba, Agric. Biol. Chem. 42 (1978) 763.
a) L. D. Stegink, J. A. Shepherd. L. K. Fry, L. J. Filer, Pediutr Rex 8
(1974) 396; b) B. Huber, H. KrPmer, F. Led], L. Vogel. Krankenhausphurm a z e i0 (1989) 168.
a) J. Beck. F. Ledl, T. Severin, Carbohydr. Rex f 7 7 (1988) 240; b) Z .
Lehensm. Unters. Forsch. 188 (1989) 118; c) H. Huber, F. Ledl. Curhohydr. Res., in press.
E. F. L. J. Anet, Austr. J Chem. 10 (1957) 193
a) W. Baltes, K. Franke, 2. Lebensm. Unters. Forsch. i67 (1978) 403; b)
L. Kroh, G. Westphal, 2. Chem. 23 (1983) 293.
a) J. F. Carson. J Am. Chem. Soc. 77 (1955) 1881. 5957; b) K. Heyns.
R. Eichstedt, K. H. Meinecke, Chem. Ber. 88 (1955) 1551.
K. Heyns, G. Miiller, H. Paulsen, Justus Liebigs Ann. Chem. 703 (1967)
202.
W. Nedvidek, F. Ledl, unpublished.
M. S . Feather, J. F. Harris, Adv. Curbohydr. Chem. Biochem. 28 (1973)
161.
591
[19] G. Machell. G N. Richard, J Chem. Soc. 1960, 1924, 1932. 1938.
[20] a) E. F. L. J. Anet, Ausrr. J Chem. 13 (1960) 396; b) J Am. Chem. Suc. 82
(1960) 1502: c) H . Kato, Bull. Agrte. Chem. Soc. Jpn. 24 (1960) 1.
[21] a) E. F. L. J. Anet, Chem. Ind. (London) 1962, 262. b) J Chromutogr. 9
(1962) 291.
[22] K. Olsson, P. A. Pernemalm, 0. Theander, A C I UChem. Scund Ser. B 32
(1978) 249.
(231 J. Beck, E Ledl, M. Sengl, T. Severin. 2. Lehensm. Unters. Forsch. 190
(1990) 212.
[24] a ) E. F. L. J. Anet. Tetrahedron Lett. 1966, 1649; b) Carhohydr. Res 2
(1966) 448.
[25] a ) R. Kuhn, G. Kriiger, Chem. Ber. 90 (1957) 264; b) J. N. BeMiller.
R. L. Whistler, J Org. Chem. 27 (1962) 1161.
[26] G. R. Jurch, J. H. Tatum, Carhoh!dr. Re$ IS (1970) 233.
[27] 0. Pachmayr. F. Ledl, T. Severin, Z. Lebensm. Unters. Fm-sch. 182(1986)
294
[28] K. Olsson, P. A. Pernemalm. T. Popoff, 0. Theander, Actu Chem. Srand.
Ser. B 31 (1977) 469.
[29] H. Shigematsu, T. Kurata, H. Kato, M. Fujimaki, Agric. B i d . Chem. 35
(1971) 2097.
[30] F. Hayase. R. H. Nagdraj, S. Miyala, F G. Njoroge, V. M. Monnier, J
Biol. Chem. 264 (1989) 3758.
[31] a) F Ledl, M. Sengl, T. Severin, J Chromatogr. 463 (1989) 119; b)
U. Schussler, F. Ledl. 2. Lehensm. Unter.7. Forsch. 189 (1989) 32.
[32] H. Kato. Agrrc. B i d . Chem. 31 (1967) 1086.
[33] R. Tressl, B. Helak, D . Rewicki. Helv. Chim. Acta 65 (1982) 483.
[34] F. Ledl, U. Kronig, T. Severin. H. Lotter. 2. Lehensm. Unters. Forsch.
177 (1983) 267.
[35] F. G. Njoroge, A A. Fernandes. V. M. Monnier, J Carhohydr. Chem. 6
(1987) 553.
[36] J. G. Farmar, P. C. Ulrich. A. Cerami, .
I
Org. Chem. 53 (1988) 2346.
[37] a) T. Severin, U. Kronig, 2. Lehensm. Unters. Forsch. 152 (1973) 42; b)
S. Estendorfer, F. Ledl. T. Severin, unpublished: c) U . Kronig. Dfsserration, Universitit Miinchen 1974: d ) Estendorfer. F. Ledl, T. Severin,
Angew. Chem. 102 (1990) 547; Angew Chem. Int. Ed. Engl. 29 (1990) 536.
[38] J. Hiebl, F. Ledl, T. Severin, J. Agric. Food Chem. 35 (1987) 990.
[39] T. Severin. W Seilmeier, 2. Lehensm. Uflters. Forsch. 134 (1967) 230.
[40] J. E. Hodge. B. E. Fisher. E. C. Nelson, Proc. Am. Soc. Bren. Chem.
1963. 84.
[41] G. A. M. van den Ouweland. H. G. Peer, J Agric. Food Chem. 23 (1975)
501.
[42] A. 0. Pittet. P. Rittersbacher, R. Murahdhara, J Agric. Food Chem. 18
(1970) 929.
[43] a) J. 0. Rodin, C. M. Himel, R. M. Silverstein, R. W. Leeper. W. A.
Gortner, J. Food Sci. 30 (1965) 280; b) B. Willhalm, M. Stoll. A F.
Thomas, Chem. b d . (LondonJ 1965. 1629.
[44] a) T. Severin. W. Seilmeier, 2. Lehensm. Unters. Forsch. 137 (1968) 4; b)
F. D. Mills, D Weisleder. J. E. Hodge, Tetrahedron Lett. 1970, 1243.
1451 a) F. Ledl, W Schnell, T. Severin, 2. Lehensm. Unrers. Forsch. 160 (1976)
367; b) F Ledl, Lehensmittelchem. Gerichtl. Chem. 41 (1987) 83.
[46] a) F. D. Mills, B. G. Baker, J. E. Hodge, Carhohydr. Res. 15 (1970) 205:
b) H. Simon, G Heubach, Chem. Ber. 98 (1965) 3703; c) F. Ledl,
T. Severin. 2. Lehensm. Unturs. Forsch. 169 (1979) 173.
[47] F. Ledl, G. Ellrich, H. Kostermeyer, Z . Lebensm. Unters. Forsch. I82
(1986) 12.
[48] H. Matsutira, Y. Yoshida, K . Kumihiro, T. Fuwa, R. Kasai, 0. Tanaka.
Chem. Pharm. Bull. 32 (1984) 4674.
1491 J. E. Hodge, E. C. Nelson, Cereal Chem. 38 (1961) 207.
[50] J. Rothmoser, Di.ssertation. Universitit Miinchen 1981.
[51] a) J. Brand, Ber. Dtsch. Chem. Ges. 37 (1894) 807: b) S. Patton. J Eiol.
Chem. 184 (1950) 131.
[52] J C. Goodwin, Curhohydr. Res. 115 (1983) 281.
[53] F. Weygand, H. Simon. W. Bitterlich, J. E. Hodge, B. E. Fisher, Tetrahedron 6 (1958) 123.
[54] F. Ledl, G. Fritsch, 2. Lebensm. Unlers. Forsc,h. 178 (1984) 41.
[ 5 5 ] F. Ledl, 2. Lebensm. Unters. Forsch. 179 (1984) 381.
1561 T. Severin. A. Loidl, Z. Lehensm. Unrrrv. Forsch. 161 (1976) 119.
[57] F. Ledl, H . Osiander. 0. Pachmayr, T. Severin, 2 Lebensm. Unters.
Forsch. 188 (1989) 207.
1581 G. J. Kontoghiorghes, R. W. Evans, FEES Lett. 189 (1985) 141.
[59] W. 0. Nelson, S J. Rettig, C. Orvig, J. Am. Chem. Soc. 109 (1987) 4121
[60] H . G. Peer, G. A. M. van den Ouweland, C N. De Groot. Reel. Tray.
Chim. P q s - B a s 87 (1968) 1011.
[61] T. Doornbos, G. A. M. van den Ouweland, S. B. Tjan in [7c], p. 57.
[62] F. Ledl, T. Severin, 2. Lebensrrr. Unrers. Forsch. 167 (1978) 410.
[63] R. E Miller, S. M. Cantor, J Am. Chem. Soc. 74 (1952) 5236.
(641 a) K . Heyns, J. Heukeshoven, K. H. Brose, Angew. Chem. RO (1968) 62,
Angew. Chem. Int. Ed. Engl. 7 (1968) 628: b) P. A. Finot, J. Bricout.
R. Viani, J. Mauron, Experientra 24 (1968) 1097; c) H. Erbersdobler.
H. Zucker. Milchnissen.rcha/t 21 (1966) 564.
[65] a) B Huber. F. Ledl. T. Severin, A. Stangl, G. Pfleiderer. Curhohydr. Res.
I82 (1988) 301: b) F. G. Njoroge, A. A. Fernandes, V. M. Monnier, J
Bfol. Chem. 263 (1988) 10646.
592
[66] G. Westphal, L. Kroh, E. Sandner, Nahrung 25 (1981) 893.
[67] T. Hayashi. M. Namiki in [7g], p. 29.
[68] a) F. Hayase, H . Kato in [7g], p. 39; b) T. Severin, J. Hiebl, H. PoppGinsbach. Z. Lehensm. Unters. Forsch. 178 (1984) 284.
[69] a ) G. Manninger, T Severin, 2. Lebensm. Unrers. Forsch. 161 (1976) 45;
b) K. Heyns. U. Sage, H. Pauisen. Curhohydr. Res. 2 (1966) 328; c) M. PiIoty, W. Baltes. 2. Lehensm. Unters. Forsch. 168 (1979) 374; d) T. Severin,
T. Wieland, ihid. 153 (1973) 201: e) N. van Chuyen, T. Kurata, M. Fujimaki, Agric. B i d . Chem. 37 (1973) 1613.
[70] a) M. Namiki. T. Hayashi in [7c], p. 81; b) in [7d], p. 21; c) T. Hayashi,
S. Mase. M. Namiki, Agric. Biol. Cheni. 49 (1985) 3131.
[71] a)T. Hayashi. S. Mase, M. Namiki. ARrfc. Biol. Chem. SO(1986) 1959; b)
T. Hayashi. M. Namiki in [7g], p. 29.
I721 Y. Kato. T. Matsuda, N. Kato, K. Watanabe, R. Nakamura. J Agric.
Food Chem. 34 (1986) 351.
[73] Y. Kato, T. Matsuda, N. Kato, R. Nakamura, J Agric. Food Chem. 36
(1 988) 806.
[74] M. Sengl, F. Ledl, T. Severin, unpublished.
[75] F. HaydSe, H. Kato, Agric. Biol. Chem 49 (1985) 467.
[76] M. F. Aly, R. Grigg, Tetrahedron 23 (1988) 7271.
[77] M U. Ahmed. S. R. Thorpe, 3. W Baynes, J Bzoi. Chem. 261 (1986)
4889; J. A. Dunn, D. G. Dyer, K. J. Knecht, S. R. Thorpe, D . R.
McCance. K. Bailie. G. Silvestri, T. J. Lyons, J. W. Baynes in [7k]. p. 425.
1781 R. Liardon. D. de Weck-Gaudard. G. Philippossran, P. A. Finot, J
Agric. Food Chem. 35 (1987) 427.
[79] A. Schonberg, R. Moubacher, Chem. Rev. 50 (1952) 261.
[SO] R. Miller. K. Oisson, P A. Pernemalm, Acta Chrm. Scand. Ser. B 38
(1 984) 689.
[Xl] A. F. Gluron, B. Quack, T. P. Mawhinney, M. S. Feather, J Agric. Food
Chem. 36 (1988) 677; W. Baltes. L. Mevissen, 2. Lebensm. Unrers. Forsch.
I87 (1988) 209.
[82] M. P. Piloty, W. Baltes, 2. Lehensm. Unters. Forsch. 168 (1979) 368, 374.
[83] P. E. Shaw. R. E. Berry, J Agric. Food Chem. 25 (1977) 641
[84] T. Nyhammar. K. Olsson. P. A. Pernemalm, ACIUScanrl. Chem. Ser. B 37
(1983) 879.
[85] M. Namiki. A. Terao, S. Ueda, T. Hayashi, Agric. Biol. Chem. 49 (1986)
3 139.
[X6] a) H. K. Nielsen, D. de Weck, P. A. Finot, R. Liardon, R. F. Hurrell, Br.
J Nutr. 53 (1985) 281 ; b) F. Hayase, H. Kato. M. Fujimaki, Agric. Biol.
Chem. 43 (1979) 2459.
[87] S. Nakasuka, B. Feng, T. Goto, K . Kihara, Tetrahedron Lett. 27 (1986)
3399.
(881 a ) T. Severin, B. Heidenhain, Z. Lebensm. Unters. Forsch. 132 (1966) 65;
b) T. Severin, K. H. Briiutigam, Chem. Ber. 106 (1973) 2943; c)
K. H . Briutigam. T. Severin. 2. Lehensm. Unters. Forsch. 154 (1974) 80:
d) E. Knoch, W. Baltes, Lehmsmittelchem. Cerichtl. Chem. 42 (1988) 110;
e) T. R. Bosin, S. Krogh, D. Mais, J Agric. Food Chem. 34 (1986) 843.
[89] T. Nyhammar, P. A. Pernemalm, Food Chem. 17 (1985) 289.
[90] a) T. Sugimura, T. Kawachi, M. Gajao, T. Gakagi, Y. Seino, T. Okamoto,
K. Shodo, T. Kosuge, K. Tsuji, K. Watabayashi, Y. litaka, A. Itai, Proc.
Jpn. Acad. 53 (1977) 58; b) M. Nagao, S. Sato, T. Sugimura in [7d].
p. 521.
1911 2. Yamaimmi. T. Shiomi. H. Kasai. S. Nishimura, Y. Takahashi. M. Nagao, T. Sugimura, Cancer Lett. (Shannon Irel.) 9 (1980) 75.
[92] a) H. Shigematsu. S. Shibdta. T. Kurata, H. Kato. M. Fujimaki. J Agric.
Food Chem. 23 (1975) 233: b) R. Tressl, B. Helak, H. Kamperschroer,
N. Martin, D. Rewicki, ihid. 33 (1985) 919: c) B. Helak, E. Kersten,
K. Sprengler. R. Tressl, D. Rewicki, ihid. 33 (1985) 919; c) ihid 37 (1989)
405.
[93] a) R. Tressl, D. Rewicki, H. Kamperschroer, B. Helak, J Agric. Food
Chem. 33 (1985) 924; b) B. Helak, K . Sprengler, R. Tressl, D. Rewicki,
ibrd. 37 (1989) 400.
[94] a) R. Tressl, B. Helak. K. Sprengler. A. Schroder. D. Rewicki, Liebigs
Ann. Chem. 1985,2017:h) R. Tressl, B. Helak, D. Rewicki, J Agric. Food
Chem. 33 (1985) 1132; c) H. M. E. Pabst, F. Ledl. H.-D. Belitz. Z.
Lehensm. Unters. Forsch. 181 (1985) 386.
[95] a) R. Tressl. K. G. Griinewald, E. Kersten, D. Rewicki, J. Agrfr. Food
Chem. 34 (1986) 347; h) /hid. 33 (1985) 1137.
[96] a) R. A. Wilson, 1. Katz. Flavour Ind. 1974. 30, b) G. MacLeod in G. G.
Birch, M. G. Lindley (eds.): Developments in Food Flavours, Elsevier
Applied Science, London. New York 1986, p. 191, c) M. J. Lane,
H. E. Nursten in [7d], p. 141
[97] a) E. J. Mulders. 2 Lebensm. Unters. Forsch. 152 (1973) 193; b) F. Ledl,
T. Severin, Chem. Mikrohiol. Techno1 Lebensm. 2 (1973) 155: c) F. Ledl.
T. Severin, Z. Lehensm. Unters. Forsch 154 (1974) 29.
[98] a) 0. G. Vitzthum, P. Werkhoff, J FoodSci. 39 (1974) 1210; b) R. Tressl
in [7j],p. 285.
1991 R. A. Wilson, C. J. Mussinan, I. Katz. A. Sanderson. J Agric. Food
Chem. 21 (1973) 873.
[loo] a) R. K. Cho. A. Okitani, H. Kato in [7g], p.439: b) D . B. Shin,
F. Hayase, H. Kato, Agric. B i d . Chem. 52 (1988) 1451
[lo11 D. R. Sell. V. M. Monnier, J. Biol. Chem. 264 (1989) 21597.
[lo21 K. Heyns. H. Noack, Chenz. Ber. 95 (1962) 720.
Angew. Chem. Int. Ed. Engl. 29 (1990) 565-594
11031 E.Klein. B. Kramholler, F. Ledl, T. Severin, unpublished.
[lo41 T. Nyhammar, S. Grivds, K. Olsson, M. Jigerstad in [7g], p. 323.
[lo51 D. Yosida. Y. Saito, G. Mizugdki, Agric. Biol. Chem. 48 (1984) 241.
[lo61 a) H. Kasai, Z. Yamaimmi. K. Wakabayashi, M. Nagao, T. Sugimura,
S. Yokoyama, T. Miyazawa. S. Nishimura, Chem. Lett. 1980, 1391; b)
H. Kasai, 2. Yamaimmi, T. Shimoi, S. Yokoyama, T. Miyazawa,
K . Wdkabayashi. M. Ndgao, T. Sugimura, S. Nishimurd, ihid. 1981,485;
c) K. Wdkabayashi, M. Takahashi, M. Nagao. S. Sato, N. Kinae, I.
Tomita, T. Sugimura in [7g], p. 363.
[lo71 R. C. Jones, J. H. Weisburger, Mutar. Res. 206 (1988) 343; b) Jpn. J Canci'r Rrs. iCannJ 79 (1988) 222
[lo81 K. Aso. Nippon Nogei Kaguku Kaishi 16 (1939) 249.
I1091 a ) H. Tsuchida. M. Komoto, K. Kato, M. Fujimaki, Agric. Biol. Chem.
37 (1973) 2571; b) H. Tsuchida, K. Morinaka, S. Fujii, M. Komoto,
S. Mizuno in [7gl, p 85; c) H. Tsuchida, S. Tachibana, M. Komoto,
Agric. Biol. Chem. 40 (1976) 1241; d) M. Komoto, H. Tsuchida. ihid. 44
(1980) 11x9.
Ill01 a) U . Kroplien, J. Rosdorfer, J. Van der Greef. R. C. Long, J. H.
Goldstein. J Org. Chem. 50 (1985) 1131; b) S. Fujii, H. Tsuchida, M.
Komoto. Agric. Biol. Chem. 30 (1966) 73.
[I 111 H. Tsuchida, K. Kitamura, M. Komoto, Carbohvdr. Res. 67 (1978) 549.
[112] W. Seibel. J. M. Briimmer, Dfsch. Lehensm. Rundsch. 83 (1987) 171.
[113] a) K. Nishie. A. C. Weiss, A. C. Keyl, Toxrcol. Appl. Pharmucol. 14
(1969) 301; b) I. F. Gaunt, A. G. Lloyd, P. Grasso. S. D. Gangolli,
K. R. Butterworth, Food Cosmet. Toxicol. 15 (1977) 509.
I1141 a) J. F Lawrence, C. F. Charbonneau, J Chromafogr.407 (1987) 405; b)
U. Kroplien. ihid. 362 (1986) 286.
11151 a) R. Kuhn. G. Kriiger, H. J. Haas, A. Seliger. Justus Liebigs Ann. Chem.
644 (1961) 122; b) G. Candiano, G. M. Ghiggeri, R. Gusmano. L. Zettd,
E. Benfenati, G. Icardi. Curhohydr. Res. 184 (1988) 67.
[116] a ) H. Motai. S. Inone, Agric. Biol. Chem. 38 (1974) 233; b) S. Homma,
T. Tomurd, M. Fujimaki, ihid. 46 (1982) 1791
[117] K Olsson, P. A. Pernemalm, 0. Theander in [7c], p. 47.
[IlX] a)H.Kato,S.B.Kim,F.Hayasein[7g],p.215;b)M.S.Feather,R.D.Huang in [7gJ. p. 183; c) L. M. Benzing-Purdie, C. I. Ratcliffe in [7g],
p. 193; d) R. D. Huang. M. S. Feather, J. Agrir Food Chem. 36 (1988)
673.
[I 191 at S. B. Kim, F. Hayase, H. Kato, Agric. Biol. Chem. 49 (1985) 785. b)
F. Hayase. S. B. Kim. H. Kato, ihid. 48 (1984) 2711.
[120] a ) F Hayase. H. Kato. Agric. B i d . Chem. 45 (1981) 2559; b) F. Hayase,
M Sato. H. Tsuchida, H. Kato, ihid. 46 (1982) 2987; c) H. Tsuchida.
M. Komoto. H. Kato, M. Fujimaki, rhid. 40 (1976) 2051; d) K. Heyns,
R. Hauber. Justus Liehlgs Ann. Chem. 733 (1970) 159; e) W. Bakes,
K Franke, W Hortig, R. Otto. U. Lessig in [7c], p. 137; f) H. Kato,
H. Tsuchida in [7c], p. 147.
11211 a ) R. Hardt. W. Baltes, Z.Lebensm. Linters. Forsrh. 185 (1987) 275; b)
A Dross, R. Hardt, W. Baltes, Fresenius Z . Anal. Chem. 328 (1987) 495.
[122] C . H. Wu. G. Russell, W. D. Powrie. J FoodSci. 52 (1987) 813.
[123] M. Hermann. F. Ledl, unpublished.
[124] a ) S Forsin[7d],p. 185;b)G. Ohloff,I. Flament,He~erocyclesll(1978)
663; c ) W. Baltes, Dtsch. Lebensm. Rundsch. 75 (1979) 2, d)
H E. Nursten, Food Chem. 6 (1981) 263; e) J. P. Danehy, Adv. Food Res.
30 ( 1 986) 77.
I1251 H. W. Brinkman, H. Copier, J. J. M. de Leuw. S. B. Tjan, J. Agric. Food
Chcm. 20 (1972) 177.
[126] U Gasser. W. Grosch, 2. Lehensm. (infers. Forsch. 186 (1988) 489.
[1271 a ) P. Schieberle, W. Grosch, 2. Lehensm. (inters. Forsrh. 177 (1983) 173;
ihrd. 178 (1984) 479; ihid. 180 (1985) 474; b) J. Agric. Food Chem. 35
(1 9x7) 252.
[1281 a ) I . Flament. B. Willhalm, M. Stoll. Helv. Chim. Acts 50 (1967) 2233; b)
R. G. Buttery, R. M. Seifert, D. G. Guadagni, L. C. Ling, J: Agric. Food
Chem. 19 (1971) 969; c) C. A. Pittet, D. Hruza, ihd. 22 (1974) 264.
[1291 P. Schieberle. in [7j]. p. 350 and in [7k], p. 184.
11301 C. Wan-Chung, Dissertation, Technische Universitit Miinchen 1979.
[131] a ) H. M. E. Pabst, F. Ledl, H.-D. Belitz, 2. Lehensm. Linters. Forsch. 178
(1984) 356; b) H. M. E Pabst, H. Lotter. H.-D. Belitz, ihid. /86 (1988)
311.
[132] W. Baltes, G. Bochmann, J. Agric. Food Chem. 35 (1987) 340.
[133] F. Ledl. J. Hiebl, T. Severin, 2. Lehensm. Unters. Forsch. 177 (1983) 353.
1134) F. Ledl. H. Lotter, unpublished.
[135] F. Ledl, T. Severin, 2. Lehensm (inters. Forsch. 175 (1982) 262.
11361 a ) T. Severin. U. Kronig, Chem. Mikrohiul. Technol. Lebmsm. 1 (1972)
153; b) U. Kronig, T. Severin, &id. 2 (1973) 49; c) H. E. Nursten,
R. O'Reilly in [7d], p. 103.
[137] H. Hashiba in [7g]. p. 155.
[1381 a)T. Kurata, M. Fujimaki, Y. Sakurai, Agric. Biol. Chem.37(1973) 1471;
b) T. Hayashi, Y. Hoshi, M. Namiki, i h d 47(1983) 1003; ihid. 48 (1984)
2377; c) T. Kurdta. T. Imai, N. Ardkawa in [7g], p. 67
[1391 H.Sakurai. K. Ishii, Bull. Coll. Agric. Vet. Mrd. Nrhon (iniv. 45(1988) 50,
59.
11401 B. Huber. P. Noll. F. Ledl, unpublished.
11411 H. Pehr. G. Manninger, T. Severin, Curhohvdr. Res. 64 (1978) 1.
[1421 L. Mevissen. W. Baltes, Z. Lebensm. (inters. Forsch. 176 (1983) 206
Angew Chcm. 102 (19901 S6S-594
[143] P. Noll, F. Ledl, unpublished.
ti441 P. E.Shaw, J. H. Tatum, R. E. Berry, Carhohydr. Res. 16 (1971) 207.
(1451 F. Ledl, G. Fritsch, T. Severin, Z. Lebensm. Linters. Forsch. 175 (1982)
208.
[146] F. D. Mills, J. E. Hodge. W. K. Rohwedder, L. W. Tjarks, J. Org. Chem.
38 (1973) 2512.
[147] J. E. Hodge, C. D. Evans, US-Pat. 2806794 (1957).
[148] R Schnee. K. Eichner, 2. Lebensm. (inters. Forsch. 185 (1987) 188.
I1491 F. Ledl, G. Fritsch, J. Hiebl, T. Severin. 2. Lehensm. (infer.\. Forsch. 176
(1983) 294.
[I501 a) H. Lingnert, C. E. Eriksson in [7d], p. 335; b) N. Yamdguchi in [7g],
p. 291.
11511 B. Commoner, A. J. Vitayathil, P. Dolard, G. Nair, P. Madyastha,
G. C Cuca, Science (Washington D . C . ) 201 (1978) 913.
[152] M. Scheutwinkel-Reich. G. Reese, W. van der Hude, Z. Lehensm. Clnrors.
Forsch. 180 (1985) 207.
[153] a) M. Jigerstadt, A. L. Reuterswird, R. h e . M. Dahlquist in [7d],
p. 507; b) K. Shinohara. R. T. Wu. N. Jahan. M. Fdnaka, N. Moringa,
H. Murakami. H. Omura, Agric. Biol. Chem. 44 (1980) 671.
[154] a) H. Ohgaki, K. Kusdma. N. Matsukura, K. Marino. H. Hasegawa,
S. Sato, S. Takayama, T. Sugimura, Carcinogenesis (London) 5 (1984)
921; b) H. Ohgaki, H. HasegdWa, T. Kato, M. Suenaga. M. Ubutaka,
S. Sato, S. Takayama, T. Sugimura, Proc. Jpn. Acad. Ser. B 61 (1985)
137; T. Sugimurd, S. Takayama, H. Ohgaki, K . Wakabayashi, M. Nagao
in [7k], p. 323.
[155] a) S. B. Kim, F. Hayase, H. Kato in [7g], p. 383; b) T. Kada. K. Moritd,
T. Inone, M u m . Res. 53 (1978) 351; c) R. I. M. Chan, H. F. Stich,
M. P. Rosin, W. D. Powrie, Cancer Lett. (Shannon Ire/.) I S (1982) 27.
[156] a) D. W Gruenwedel, S. C. Lynch, G. F. Russell, Adv. E.rp. Med. Biol.
177 (1984) 269; b) K. Heyns, S. Roper. H. Roper, B. Meyer, Angrw.
Chem. 91 (1979) 940; Angew. Chem. Int. Ed. Engl. 18 (1979) 878.
11571 N. Kinae in [7g]. p. 343.
[158] S. Mihara, T. Shibamoto, J: Agric. Fuod Chem. 28 (1980) 62.
11591 A. Nagahara, K. Ohshita. S. Nasuno in [7gI, p. 373.
(1601 H. U. Aeschbacher in [7k]. p. 335.
[161] a ) H. Omura, N. Jahan. K. Shimohara, H. Murakami in [7d], p. 537; b)
M. Ndgao, T. Yahagi, M. Honda, Y. Seino. T. Matsushima, T. Sugimura.
Proc. Jpn. Acud. 53 (1977) 34.
[I621 a) J. Mauron, F. Mottu, Arch. Biochem. Biophys. 77 (1958) 312; b)
P. A. Finot, J. Mauron. Helv. Chim. Acta 52 (1969) 1488; c) 1. Mauron.
I n t . J. Vifam. Nufr. Rrs. 40 (1970) 209; d) P. A. Finot, E.Magnenat.
F. Mottu, E. Bujard. Ann. Nurr. Aliment. 32 (1978) 325; e) B. Sherr,
C. M. Lee, C. Jelesciewicz, J. Agrrc. Food Chem. 37 (1989) 119
O/.
[1631 D. E. Furniss, R. F. Hurrell. P. A. Finot. Arra Pharmaco/. R J X ~ C59
(1986) 188.
[164] P. A. Finot. D E. Furniss in [7i], p. 343.
[165] M. Fujimaki. S. Homma, N. Ardkawa, C. Inagdki. Agrrc. Bwi. Chem. 43
(1979) 497.
[166] G. J. Pintauro, T C. Lee, C. 0. Chichester in [7d], p. 467.
[167] a) H. S. Burton, D. J. McWeeny, D. 0. Biltcliffe, J. Scr. Food Agrfc. 14
(1963) 911; b) B. L. Wedzicha (Ed.): Chemistry of' Sulfur Dioxide in
Foods. Elsevier Applied Science, London, New York 1984.
[168] H. E Nursten in D. MacCarthy (Eds.): Concentration and Drying of
Foods, Elsevier Applied Science, London, New York 1986, p. 53.
[169] a) M. E. Knowles, Chem. Ind. 1971, 910; 8 . Lindberg, J. Tdnaka, 0 .
Theander, Acra Chem. Scand. 18 (1964) 1164; E. F. L. J. Anet, D. L
Ingles, Chem. Ind. /964. 1319. b) B. L. Wedzicha, M. T. Kaputo. J Food
Scr. Trchnol. 22 (1987) 643.
[170] a) R. G. Arnold, J. Dairy Sci. 52 (1969) 1857; b) L. Kroh. S. Zeisc.
G Westphal. R. Mdcholz, Nahrung 33 (1989) 5 1 ; c) L. Kroh, S. Zeisc,
R. Stosser, G. Westphal. 2. Lehensm. (inter.?. Forsch. 188 (19x9) 115.
[171] D. W. Allen, W. A. Schroeder, J. Balog, J: A m . Chem. Soc. 80(1958) 1628.
[172] a) H. F. Bunn, D. N. Haney, K. H Gabbay, P. M. Gallop, Biochenr.
Biophys. Res. Comm. 67 (1975) 103. H. F. Bum, R. Shdpiro, M.
McManns. L. Garrick, M. J. McDonald, K. H. Gallop. J. Bio/ C/iem
254 (1979) 3892; b) A. L. Trivelli, H. M. Rdnney, H. T. Ldi, N. Engl. J
Med. 284 (1971) 353; see also Ref. 11741
[173] S. Rahbar. 0. Blumenfeld, H. M. Ranney. Biochem. Biophys. Res. Commun. 36 (1969) 838.
11741 H. F. Bunn, D. N. Haney, K. H. Gabbay. P M. Gallop, Biochem. Biophys. Res Commun. 67 (1975) 103.
[175] a) R. Fluckiger, K. H. Winterhalter, FEBS Left. 71 (1976) 356; b)
R. Dolhofer. A. StZdele, 0. H. Wieland. Klin. Wochenschr. 55 (1977)
945.
[176] a) J. W. Baynes, S. R. Thorpe, M. M. Murtiashaw, Methods Enz.vmo1.
f06 (1984) 88; b) E. Schleicher, 0. H. Wieland, Biochim. Bi0ph.v.s. Acra
884 (1986) 199.
[177] H. E. Schultze, J. F. Heremans (Eds.): Molecular Bio1og.v u/ Human
Proteins, Elsevier. Amsterdam 1966. p. 473.
[178] A glucoprotein from Hulobacterrum halobium has been described which
contains glucose bound N-glycosidically to the amide nitrogen of the
asparagine. F. Wieland, R. Heitzer, W Schdfer, Proc. Narl. A m d . Sci.
U S A 80 (1983) 5470.
593
[179] Fructose also reacts to a certain extent with proteins in vivo to give
2-amino-2-deoxyglucose 2 (Heyns-product): J. D. McPherson, B. H.
Shilton, D. J. Walton, Biochemi.7fry 27 (1988) 1901.
I1801 C. I. Neglia, H. J. Cohen. A. R. Garber, P. D. Ellis. S. R. Thorpe, J. WBaynes, J. Bio/. Chem. 258 (1983) 14 279.
[181] E. Schleicher. 0 . H. Wiekand, J. Clin. Chem. Clin. Biochem. I9 (1981) Xi.
[182] 0. H. Wieland. R. Dolhofer, E. Schleicher, Int. Congr. Ser. Excerptu
Med. n 500 (1980) 721.
[183] a) R. Dolhofer, 0. H. Wieland. Diabetes 29 (1980) 417; b) Clin. Chim.
Actu 112 (1981) 197.
[184] R. N. Johnson, P. A. Metcalf, J. R. Baker, Clin.Chim. Acru 127 (1983)
87
[185] a) R. Fliickiger, P. M. Gallop, Methods Enzymol. 106 (1984) 77, b) K.
Miedemd, T. Caspdrie, Ann. Clin. Biochem. 2 / (1984); b) E. Schleicher,
Internist 26 (1985) 372.
(1861 R. W. Fischer, K. H. Winterhalter, FEBS Letf. 135 (1981) 145.
[I871 L. K. Curtiss, J. L. Witztum, J. Clin. Invest. 72 (1983) 1427
[lXSj M. Brownlee. H. Vlassara, A. Cerami, Diabetes 29 (1980) 1044.
[189] E. Schleicher, R. Maier, unpuhlished.
[190] a) J. F. Day, S. R. Thorpe, J. W Baynes, J Biol.Chem. 254 (1979) 595; b)
R. Dolhofer, 0. H. Wieland, FEBS Lett. 103 (1979) 282.
[191j a) M. Cohen (Ed.). Diabetes and Protein Glycosylution, Springer, Berlin
1986; b) E. D. Schleicher, T Deufel, 0 H. Wieland, FEBS Left. 12Y
(1981) 1 ; c) E. D. Schleicher, K. D. Gerbitz. R. Dolhofer. E. Reindel,
0. H. Wieland, E. Edelmann, M. Haslbeck. W. Kemmler, H. Walter.
H. Mehnert, Diabetes Cure 7 (1984) 548.
[192] E. Schleicher, R. Scheller, 0. H. Wieland, Biochem. Bioph,vs. Res. Cornmun. 99(1981) 1011.
11931 B. W. Vogt, E. D. Schleicher, 0. H. Wieland, Diabetes 31 (1982) 1123.
[194] S. L. Schnider, R. R. Kohn, J. Clin. Invest. 67 (1981) 1630.
[195] R. L. Garlick, H. F. Bunn, R. Spiro, Diubetes 37 (1988) 1/44.
[196] E Reindl, Disserration, Universitat Miinchen 1983.
[197] V. M. Monnier, R. R. Kohn, A. Cerami, Proc. Null. Acud. Sci. U S A 81
(1984) 583.
[198] S. Pongor, P. C. Ulrich. F. A. Bencsath, A. Cerami, Proc. Nut/. Acad.
Sci. U S A 81 (1984) 2684.
594
[199] T. J. Lyons, G. Silvestri, J. A. Dunn, J. W. Baynes, Conference on “Glycated Proteins in Diabetes mellitus”, Adelaide, Australia 1988.
[200] A. Cerami. H. Vlassara, M . Brownlee. Diabetes Cure / f Suppl. 1 (1988)
73.
[201] R. Bucala. P. Model. A. Cerami, Proc. Nut[. Acud. Sci. U S A 81 (1984)
105.
[202] V. J Stevens, C. A. Rower, V. A. Monnier, A. Cerami, Proc. Null. Acud.
Sci. U S A 75 (1 978) 2918.
[203] a)S. H. Chlou. L. T. Chylack, H. F. Bum, J. H. Kinoshita, Biochem. Biophys. Res. Commun. 95 (1980) 894; b) V. M. Monnier, A. Cerami, Science
(Wushingfon D.C.) 211 (1981)491.
[204] a) D. E. Harrison, J. R. Archer, Exp. Geronml. 13 (1978) 75; b)
M. J. C. Kent. N. D. Light, A. J. Bailey, Biochem. J. 225 (1985) 745.
[205] A. S. Eble, S. R. Thorpe, I W. Baynes, J. Bid. Chem. 258 (1983) 9406.
[206] M. Brownlee, H. Vlassara, A. Cerami, Diahetes 34 (1985) 938.
12071 H. Vlassara, M. Brownlee, A. Cerami, Clin. Chem. iWinston-Salem
N.C.) 32 (1986) 37.
[2081 R. Dolhofer, E. A. Siess, 0 . H. Wieland, Hoppe-SeylerS 2. Physml.
Chem. 363 (1982) 1427.
12091 a ) M. Brownlee, H. Vlassara, A. Cerami, Diabetes32 (1983) 680; b) Diubetologiu 33 (1984) 532.
[210] R. Dolhofer, 0. H. Wieland, FEBS Lett. l(J0 (1979) 133.
[211] a ) J. L. Witztum, E. H. Mahoney, M. J. Branks. M. Fisher, R. Elam,
D. Steinberg, Diabefes 31 (1982) 283; b) E. D. Schleicher, B. Olgemoller.
J. Schon. T. Durst. 0. H. Wieland, Biochim. Biophvs. Acta 846 (1985)
226.
[212] H. Vlassara. H. Brownlee. A. Cerami, J. Exp. Med. 160 (1984) 197.
12131 a) M. Oimoni, H Hatanaka, K. Ishikawa, S. Kubota, Y. Yoshimura.
S. Baba, Klin. Wuchenschr. 6 2 (1984) 477; b) M. Oimoni, S. Nishimoto,
T. Kitamura, S. Matsumoto, H. Hdtandka, K. Ishikawa, S. Baba, ihid.
6 3 (1 985) 728.
[214] V. M. Monnier, V. Vishwanath. K. F Frank, A. E. Graig, P. Dauchot,
R. R. Kohn, N. Eng. J. Med. 314 (1986) 403.
12151 M. Brownlee, H. Vlassara, A. Kooney, P. Ulrich, A. Cerami, Science
(Washington D.C. J 232 (1986) 1629.
12161 M. Wissfeld, P. Noll, F. Ledl, unpublished.
Angen. Chem. Int. Ed. Engl. 29 (1990) S6S-594
Документ
Категория
Без категории
Просмотров
0
Размер файла
2 988 Кб
Теги
reaction, aspects, maillard, body, food, human, new
1/--страниц
Пожаловаться на содержимое документа