Glycosylation of lysine-containing pentapeptides by glucuronic acid: new insights into the Maillard reaction

Article abstract of DOI:10.1016/j.carres.2009.11.031

The formation of glycosylation products in model systems consisting of d-glucuronic acid (GlcA) and lysine-containing peptides, such as Lys-Gly-Gly-Phe-Leu (1), Gly-Lys-Gly-Phe-Leu (4) and Ac-Gly-Lys-Gly-Phe-Leu (6), was examined to evaluate the site spec

Full text of DOI:10.1016/j.carres.2009.11.031

Carbohydrate Research 345 (2010) 377–384
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Glycosylation of lysine-containing pentapeptides by glucuronic acid: new
insights into the Maillard reaction
*
ˇ ´
Štefica Horvat , Maja Rošcic
-
´
Division of Organic Chemistry and Biochemistry, Ruder Boškovic Institute, POB 180, Zagreb, HR-10002, Croatia
a r t i c l e i n f o
a b s t r a c t
Article history:
The formation of glycosylation products in model systems consisting of D-glucuronic acid (GlcA) and
Received 9 November 2009
Received in revised form 25 November 2009
Accepted 30 November 2009
Available online 4 December 2009
lysine-containing peptides, such as Lys-Gly-Gly-Phe-Leu (1), Gly-Lys-Gly-Phe-Leu (4) and Ac-Gly-Lys-
Gly-Phe-Leu (6), was examined to evaluate the site specificity as well as the extent and nature of the mod-
ification. Peptides were reacted with GlcA either in solution or under dry-heating conditions. From the
incubations performed in solution (MeOH), the corresponding (1-deoxy-D-fructofuranos-1-yluronic
acid)–peptide derivatives (Amadori compounds) were isolated. Whereas reaction of 1 resulted in the for-
mation of mono-glycosylated Amadori compound 2 with the sugar moiety attached to the N -amino group
Keywords:
Amadori
e
of the Lys residue and its di-glycosylated analogue 3, exposure of 4 to GlcA afforded only di-glycosylated
peptide 5. From the incubation of GlcA with Ac-Gly-Lys-Gly-Phe-Leu (6) performed under mild dry-heating
conditions (50 °C) in an environment of 75% relative humidity, besides Amadori compound 7, two new
Maillard reaction products were isolated that contained 3-hydroxypyridinium (8) and 3-hydroxy-picolinic
acid moiety (9). The mechanism for the formation of pyridinium products is discussed.
Ó 2009 Elsevier Ltd. All rights reserved.
Glucuronic acid
Glycation
3-Hydroxypyridine
Maillard
Peptide
1. Introduction
binding may occur via two different mechanisms.14 The first is
a transacylation mechanism, where a nucleophilic group (–NH2,
Nonenzymatic glycosylation (glycation) of proteins by reduc-
ing sugars (the Maillard reaction) is a common biological phe-
nomenon that has been under active investigation for many
years.1,2 Most interest has been directed towards glycosylation
of proteins by glucose and its potential role in diabetes and
aging,3 but an increasing number of studies have been conducted
with other reducing sugars, providing evidence that they may
also react with peptides or proteins through the glycosylation
pathway.4–6 In spite of its relatively low concentration compared
–OH, –SH) on a protein attacks the carbonyl group of the acyl glu-
curonide leading to the formation of an acylated protein and free
GlcA. The second is a mechanism of Schiff base formation where
condensation occurs between the GlcA aldehyde group and the
amine group of the N-terminus and/or the
e-NH2 of a lysine
residue of a protein leading to the formation of a glycosylated
protein. Both the intramolecular rearrangement of acyl glucuron-
ides and their hydrolysis are of particular importance because
they lead to re-exposure of the hemiacetal function of the glucu-
ronic acid, thus allowing sugar–protein covalent adduct formation
through a nonenzymatic glycation reaction. The investigation of
the reactivity of these electrophilic metabolites in forming the
corresponding acyl glucuronide–peptide adduct, carried out by
trapping experiments in the presence of Lys-Phe dipeptide, dem-
onstrates that the extent of Schiff base formation is proportional
to the rearrangement rate of the parent glucuronide.15 In addi-
tion, the reactivity is strongly dependent on the inherent elec-
tronic and steric properties of each specific aglycone. The imine
formed between the glucuronic acid moiety and the primary ami-
no groups of proteins could undergo Amadori rearrangement.
Although structures of such glycation products have not yet been
formally documented, previous studies have shown that GlcA, or
reactive acyl glucuronide metabolites, can bind irreversibly to
albumin, leading to the formation of fluorescent advanced glyca-
tion end products (AGEs) similar to those that are formed with
glucose and fructose.16,17
to glucose in vivo, D-glucuronic acid (GlcA) is a reducing sugar of
biological importance due to its participation in the metabolism
of many drugs and endogenous compounds.7 A major metabolic
pathway for the biotransformation of exogenous and endogenous
carboxylic acid substrates is conjugation with endogenous GlcA to
yield acyl glucuronide metabolites implicated in a wide range of
adverse drug effects.8–10 Not unreactive as previously thought,
acyl glucuronides are potentially reactive electrophilic species
which can undergo hydrolysis, intramolecular rearrangement
(isomerization via intramolecular acyl migration) and covalent
binding to proteins, both in vivo and in vitro.11–13 The covalent
* Corresponding author. Tel.: +385 1 45 71 290; fax: +385 1 46 80 195.
E-mail address: shorvat@irb.hr (Š. Horvat).
Present address: GlaxoSmithKline Research Centre Zagreb Ltd, Prilaz baruna
Filipovica 29, Zagreb 10000, Croatia.
0008-6215/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2009.11.031

ˇ
´
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
378
To gain more detailed insight into peptide/protein glycosylation
group, the apparent destabilization of the b anomer in favour
of the
furanose tautomer was observed at Gly1 residue
:b = 95:5), whereas the ketose attached to Lys2 residue estab-
lished a 77:23 ¡b equilibrium. These data indicate that interac-
tion(s) between the 1-deoxy- -fructofuranos-1-yluronic acid and
peptide moiety in Amadori compound 5 may change the relative
populations of tautomers by inducing effects that cannot be de-
fined at present without more detailed conformational analysis.
The peptide 6 was synthesized with the glycine N-terminal ami-
no group blocked with an acetyl group to prevent glycosylation at
this site in the synthetic peptide. Following incubation with GlcA
and purification of the glycosylated peptide 7 (79%) from Ac-Gly-
Lys-Gly-Phe-Leu (6), MS analysis confirmed that peptide 6 was gly-
cated with one sugar residue per peptide molecule as indicated by
an increase in the [M+H]+ ion by 177 mass units from the parent
processes under physiological conditions, model systems using
mixtures of GlcA with peptides and proteins containing selected
structural elements should be studied. In our previous study,18
we characterized the glycosylation products generated from GlcA
and endogenous opioid peptides (enkephalins) that did not contain
lysine residues. Herein we focus on the small lysine-containing
peptides, Lys-Gly-Gly-Phe-Leu (1), Gly-Lys-Gly-Phe-Leu (4) and
Ac-Gly-Lys-Gly-Phe-Leu (6) as model peptides to examine their
susceptibility to nonenzymatic glycation reaction and to character-
ize the products formed after being exposed to glucuronic acid.
a
(a
a
D
2. Results and discussion
2.1. Synthesis and structure determination of Amadori
products formed from glucuronic acid in MeOH
peptide arising from 1-deoxy-D-fructofuranos-1-yluronic acid,
and indicating that all carbons of the original sugar backbone
(GlcA) are retained in Amadori compound 7. The spectroscopic re-
sults, derived from experiments in DMSO solution, indicate that
Systems containing GlcA and Lys-Gly-Gly-Phe-Leu (1), Gly-Lys-
Gly-Phe-Leu (4) or Ac-Gly-Lys-Gly-Phe-Leu (6) were reacted in
MeOH for three days at 50 °C. In all reaction systems, the corre-
e
N -Lys2 amino group glycosylation of peptide 6 resulted in a simi-
lar distribution of tautomers at equilibrium as observed for ketose
sponding 1-deoxy-D-fructofuranos-1-yluronic acid derivatives
attached to the Lys2 residue in compound 5 (7;
a:b = 69:31). The
(Amadori compounds) (Scheme 1) were generated. The glycosyla-
tion products were isolated from the incubation mixtures by using
preparative RP-HPLC. The structures of Amadori compounds were
confirmed by MS and NMR spectroscopy. 1H and 13C resonances
were assigned by homonuclear COSY experiments combined with
1H–13C correlation techniques through one-bond (HMQC) and mul-
tiple-bond (HMBC) 2D NMR methods.
chemical shifts for tautomers of Amadori compound 7 are listed
in Table 1.
2.2. Synthesis and structure determination of products formed
from glucuronic acid and peptides under dry-heating
conditions
Incubation of the Lys-Gly-Gly-Phe-Leu (1) with GlcA resulted in
the formation of mono-glycosylated Amadori compound 2 (15%)
with the sugar moiety attached to the N -amino group of the Lys
The formation of products was investigated by RP-HPLC in
lyophilized model systems containing GlcA and Lys-Gly-Gly-Phe-
Leu (1), Gly-Lys-Gly-Phe-Leu (4) or Ac-Gly-Lys-Gly-Phe-Leu (6)
(15:1 molar ratio of sugar to peptide) as reactants, when exposed
to 75% relative humidity environment at 50 °C, in the dark, for five
days.
e
residue, and its di-glycosylated analogue 3 (38%). Amadori com-
pounds 2 and 3 were unseparable by RP-HPLC. The assignment of
the glycosylated lysine residue was possible following the observa-
tion of the tripled set of e-CH2 and a
-CH resonances in the Lys1 re-
gion of the 13C NMR spectrum of the obtained mixture of Amadori
compounds. Analysis of the 13C chemical shift values for Lys1 resi-
due in both glycosylated and unmodified peptide 1 shows that
In contrast to the reaction performed in MeOH, when lyophi-
lized GlcA/1 sample was kept under above conditions, generation
of the numerous new products was observed; however, no major
product was detectable in the reaction mixture even after pro-
longed periods of time. Although the concentration of the parent
peptide 1 (Lys-Gly-Gly-Phe-Leu) decreased fast in the reaction
mixture reaching a value of 3% at the end of three days heating per-
iod, the RP-HPLC analysis revealed only trace amounts of Amadori
compounds 2 and 3 present in the reaction mixture. This result, in
combination with the fact that mono-glycosylated compound 2
was very unstable, while di-glycosylated Amadori compound 3
was very stable under identical dry-heating conditions (data not
shown), indicates the rapid generation and decomposition of Ama-
dori compound 2 by 1,2- and 2,3-enolization processes leading to
reactive dicarbonyl compounds, which reacted readily with lysine
amino groups in peptide 1.
e-NH2 group was preferentially glycosylated in the model peptide
system. The population of products 2 and 3 was estimated by inte-
gration of the signal intensities of the a-CH lysine carbon atoms in
the 50–61 ppm region of the spectrum. The 13C NMR spectrum
(DMSO-d6) of the glycated product contained a complex «sugar
box» region (70–104 ppm) corresponding to peaks derived from
the attached 1-deoxy-D-fructofuranos-1-yluronic acid moieties
present in the - or b-furanose form.
a
Exposure of Gly-Lys-Gly-Phe-Leu (4) to GlcA in MeOH affor-
ded after purification di-glycosylated peptide 5 in 40% yield. A
mono-substituted derivative was not detected after three days
a
of incubation, suggesting high susceptibility of both N -Gly1
e
and N -Lys2 amino groups to glycation. While MS analysis re-
When Gly-Lys-Gly-Phe-Leu (4) was incubated with GlcA under
dry-heating conditions, peptide 4 disappeared completely during
the one-day incubation period. This was associated with the for-
mation of the mono-substituted Amadori compounds (40%) having
vealed the number of sugar moieties, their positions were
unequivocally deduced from the observed large downfield shifts
(D
d ꢀ 10 ppm) of the Gly1 and Lys2 CH2–N carbons caused by
the N-alkylation at these positions. The NMR spectra (DMSO-
d6) of Amadori compound 5 showed the presence of attached ke-
the sugar moiety attached either to N -Gly1 or to N -Lys2 amino
groups of peptide 4 and with generation of the di-substituted Ama-
dori compound 5 (13%) as confirmed by NMR analysis of the iso-
lated compounds. The concentrations of Amadori products
decreased during incubation to reach after five days 7% and 8%
for mono- and di-glycated compound, respectively. After five days,
analysis of the reaction mixture by RP-HPLC revealed the presence
of eight new peaks detected by UV absorbance at 215 nm. The
peaks were isolated and analyzed by NMR spectroscopy. NMR data
indicated that none of the isolated products was homogeneous.
a
e
tose moieties in
a- or b-furanose form. The chemical shifts for
these tautomers are summarized in Table 1. The population of
tautomers in equilibrated DMSO solution of Amadori compound
5 was estimated by integration of the signal intensities of the
anomeric carbon atoms (C-2) in the 100–105 ppm region of the
13C NMR spectra. According to integrations, the
a¡b equilibrium
was shifted to the
a-furanose side with an 87:13 preference.
When comparing the equilibrium composition of sugar moieties
attached to the Gly1 with those attached to the Lys2 amino

ˇ
´
379
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
NH-R
ε
δ
γ
δ
γ
δ'
NH2
β
β
O
O
H
N
H
N
ε
α
α
γ
δ
γ
H2N
δ
N
H
N
COOH
α
α
α
H
γ
O
O
δ'
β
β
β
O
O
H
N
α
H
N
GlcA
α
γ
2
H2N
N
H
N
COOH
δ
α
α
α
H
O
ζ
O
ε
β
NH-R
ε
δ
γ
δ
δ
ζ
ε
γ
δ'
β
β
O
O
H
H
N
α
Lys-Gly-Gly-Phe-Leu (1)
α
N
R-HN
N
N
COOH
α
α
α
H
H
O
O
β
γ
3
δ
δ
ζ
ε
δ
δ'
γ
β
δ'
O
H
N
O
H
N
γ
α
β
O
α
O
α
γ
H
GlcA
α
H
α
α
γ
N
H2N
N
H
N
H
COOH
N
α
α
R-HN
N
H
N
COOH
O
O
α
O
α
β
H
β
O
β
β
γ
γ
δ
δ
δ
δ
ζ
ε
ζ
ε
ε
NH2
5
NH-R
Gly-Lys-Gly-Phe-Leu (4)
δ
γ
δ
γ
δ'
δ'
β
β
O
H
O
O
H
N
O
H
α
H
N
α
α
GlcA
α
α
N
α
γ
N
H3C-OC-HN
N
H
N
H
COOH
H3C-OC-HN
N
H
N
H
COOH
α
α
α
α
O
O
O
O
β
β
β
β
γ
δ
γ
γ
δ
δ
δ
ζ
ζ
ε
ε
ε
ε
NH-R
NH2
7
Ac-Gly-Lys-Gly-Phe-Leu (6)
6
HOOC
HOOC
O
O
2 OH
R
=
HO
HO
HO
GlcA =
4
5
CH2
1
OH
OH
3
OH
Scheme 1. Glycosylation products derived in MeOH from lysine-containing peptides 1, 4 and 6 in the presence of
D-glucuronic acid.
However, the chemical shift differences between the peptide 4 and
the products obtained suggested a dehydration process of the su-
gar residue(s) in the Amadori compounds generated, which leads
to partially unsaturated or fully aromatized compounds.
rated into compound 9. The NMR data acquired in DMSO and
compiled in Table 2 unequivocally proved the formation of 3-
hydroxypyridinium derivative 8 and 3-hydroxypicolinic acid deriv-
ative 9 under dry-heating conditions in the GlcA/6 model system.
Hydrogen/carbon assignments were validated by COSY, HMQC
and HMBC measurements. The chemical shifts of the pyridinium
heterocycle core were in agreement with those reported for
3-hydroxypyridine. Comparison of the data listed in Table 2 for
N-peptidyl 3-hydroxypyridinium derivative 8 with the NMR data
given for peptide 6 revealed almost identical chemical shifts for
The reaction of GlcA with Ac-Gly-Lys-Gly-Phe-Leu (6) under
identical dry-heating conditions led to a multicomponent mixture
from which three major products, compounds 7–9 (Scheme 2),
were isolated in 5%, 9% and 13% yield, respectively. Novel glycation
products 8 and 9 were characterized by MS and NMR spectroscopy.
The mass spectrum of compound 8 gave an ion of m/z 641, while 9
gave an ion of m/z 685 in the positive-ion mode, showing an in-
crease of 81 (C5H5O) and 125 (C6H5O3) mass units, respectively,
from the parent peptide 6. This experiment indicated that only five
carbons from GlcA are integrated into glycosylation product 8,
while a total of six carbons from the starting sugar were incorpo-
all amino acid residues, except for those of the Lys2
e-CH2 group,
which were shifted downfield by ꢀ22 ppm and 1.7 ppm in the
13C and 1H NMR spectra, respectively. Such a pronounced low-field
shift can only be rationalized if the primary nitrogen of the lysine
residue in compound 6 is converted to a tertiary one in 8. The 1H

ˇ
´
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
380
Table 1
1H and 13C NMR spectral data (d, ppm) for the Amadori compounds 5 and 7a
Residueb Atom
5
7
Anomeric form of the keto-sugar
Anomeric form of the keto-sugar
a
-furanose at Gly1 residuec
a
-furanose at Lys2 residue b-furanose at Lys2 residue
a
-furanose at Lys2 residue b-furanose at Lys2 residue
dH
dC
dH
dC
dH
dC
dH
dC
dH
dC
Ketose
1
2
3
4
5
6
3.01/3.12
—
3.74
3.85
3.98
NO
50.28
103.41
80.11
83.54
81.85
167.08
3.06/3.12
—
3.92
4.09
4.37
NO
50.28
3.62
—
3.74
4.26
4.09
NO
51.59
3.11
—
3.92
4.09
4.37
NO
50.47
3.68
—
3.74
4.25
4.09
NO
52.30
104.24
81.41d
79.21
101.13
77.07
77.39
79.67
171.87
104.26
81.34d
79.21
101.15
77.06
77.37
79.61
172.01
81.33d
172.59
81.36d
172.74
Gly1
Lys2
a
NH
CO
3.31/3.45
NO
—
50.63
—
167.79
3.71
8.08
—
42.07
—
169.13
a
b
c
d
e
NH
CO
4.23
1.45/1.64
1.26
1.55
2.91
8.14
—
52.35
31.13
22.09
24.55e
47.83f
—
4.22
1.52/1.64
1.25
1.55
2.91
8.02
—
52.30
31.26
22.29
24.67
47.85
—
22.21
22.32
24.25
47.68
24.79e
47.89f
—
2.91
2.86
47.68
2.87
a
171.38
171.43
171.64
Gly3
Phe4
a
NH
CO
3.59/3.72
8.13
—
41.70
—
168.27
3.55/3.71
8.10
—
41.72
—
168.27
a
b
c
d
e
f
NH
CO
4.56
2.74/3.04
—
7.25
7.25
7.20
8.01
—
53.46
37.62
137.61
129.13
127.92
126.15
—
4.55
2.75/3.04
—
7.24
7.24
7.18
7.96
—
53.52
37.67
137.71
129.22
127.98
126.21
—
170.90
171.05
Leu5
a
b
c
4.23
1.55
1.55
0.85/0.91
8.29
—
50.28
39.89
24.20
21.34/22.72
—
4.22
1.52
1.64
0.85/0.90
8.27
—
50.31
39.92
24.25
21.35/22.82
—
d,d0
NH
COOH
173.70
173.85
Ac
Me
CO
1.84
—
22.39
169.71
NO = not observed.
a
In DMSO-d6 at 25 °C. Those chemical shifts that show significant differences when compared with parent peptide value are indicated in bold.
b
c
One set of resonances detected for Gly1, Gly3, Phe4 and Leu5 amino acid residues.
Keto-sugar moiety in b-furanose form at Gly1 residue present in traces.
Assignment of signals could be interchangeable.
d,e,f
and 13C chemical shift differences between the compounds 8 and 9
were found to be very small with the exception of the C-2 chemical
shift value for the 3-hydroxypiridinium moiety observed in picoli-
The results thus obtained revealed that under mild dry-heating
conditions (50 °C) and in an environment of 75% relative humidity,
the conjugation between the lysine residue in peptide 6 and glucu-
ronic acid occurred readily, modifying up to 80% of the lysine ami-
no group within 48 h. The data obtained are consistent with the
previously reported observation for lysine/sugar systems showing
increased rates of the Maillard reaction in a high relative humidity
(65.5%) environment, due to increased mobility of the molecules in
the matrix.19 The formation of Amadori product 7 in the initial
phase of the dry-heating process and almost linear increases of
the pyridinium derivatives 8 and 9 as a function of time suggested
compound 7 as their precursor. In support of this hypothesis, it was
shown that enkephalin/GlcA-derived Amadori compounds decom-
pose readily to 3-hydroxypyridinium compounds under similar
dry-heating conditions.18 Ring-substituted pyridinium compounds
were also previously obtained by reaction of either triose sugars or
glucose with lysine and arginine residues in phosphate buffer solu-
nic acid derivative 9 (
D
d ꢀ 25 ppm) (Table 2).
The general profile for the formation of pyridinium compounds
8 and 9 in the GlcA/6 model system is shown in Figure 1 as a func-
tion of time. In the initial phase of the reaction, the concentration
of the parent peptide 6 decreased in the reaction mixture, reaching
a value of 40% after a heating period of one day. Under the same
conditions, the formation of Amadori compound 7 was slow to
reach the maximum concentration (24%) after three days of incu-
bation, and then it started to decrease. As presented in Figure 1,
the exposure of peptide 6 to GlcA under dry-heating conditions re-
sulted in almost equal amounts of glycation products 8 and 9. After
five days of incubation, the relative amounts of pyridinium com-
pound 8 and picolinic acid derivative 9 reached 25% and 32%,
respectively.

ˇ
´
381
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
Table 2
Ac-Gly-Lys-Gly-Phe-Leu (6)
1H and 13C NMR spectral data (d, ppm) for the pyridinium–peptide derivatives 8 and
9a
Residue
Arylb
Atom
8
9
GlcA, 50 ºC, 75% RH
dH
dC
dH
dC
2
3
4
5
8.58
—
7.91
7.93
8.50
12.27
132.88
157.09
128.46
131.21
135.42
—
—
157.63c
157.13c
130.41
131.00
132.90
δ
γ
8.19
7.92
8.59
12.27
NO
δ'
β
O
H
N
O
H
N
α
α
6
α
γ
H3C-OC-HN
N
H
N
COOH
OH
COOH
α
α
O
H
O
161.29
42.05
β
β
Gly1
Lys2
a
NH
CO
3.69
8.06
—
42.03
3.71
8.08
—
γ
6
δ
δ
HOOC
O
2
OH
ζ
ε
169.06
169.15
ε
HO
4
5
NH
a
b
c
d
4.25
1.53/1.68
1.25
1.86
4.46
8.02
—
52.02
31.08
21.66
30.14
60.52
4.24
1.52/1.69
1.32
1.84
4.79
8.04
—
52.20
31.03
22.04
30.57
59.01
CH2
1
3
OH
7
δ
e
NH
δ'
a
γ
β
O
H
N
O
H
α
CO
171.59
41.68
171.72
41.79
α
α
N
H3C-OC-HN
N
N
COOH
Gly3
Phe4
a
NH
CO
3.57/3.72
8.16
—
3.59/3.70
8.21
—
α
α
H
O
H
O
β
β
γ
γ
168.26
168.32
δ
δ
a
b
c
d
e
f
NH
CO
4.56
2.76/3.04
—
7.24
7.24
7.18
7.95
—
53.49
37.68
137.69
129.24
127.99
126.22
4.56
2.76/3.04
—
7.24
7.24
7.17
7.96
—
53.56
37.66
137.73
129.25
128.01
126.23
ζ
ε
ε
+
N
8
HO
δ
γ
171.06
171.06
δ
Leu5
a
b
c
4.22
1.53
1.63
0.84/0.89
12.60
NO
50.32
39.93
24.25
21.34/22.81
4.22
1.63
1.63
0.84/0.89
8.26
50.34
39.87
24.26
21.37/22.83
'
β
O
O
H
N
H
N
α
α
α
N
H
N
COOH
H3C-OC-HN
α
α
d,d0
NH
COOH
O
H
O
β
β
γ
γ
173.85
NO
173.86
δ
δ
Ac
Me
CO
1.83
—
22.40
169.67
1.84
22.41
169.72
ζ
ε
ε
+
HOOC
HO
N
NO = not observed.
9
a
In DMSO-d6 at 25 °C. Those chemical shifts that show significant differences
when compared with parent peptide value are indicated in bold.
b
Aryl = 3-hydroxypyridinium residue for compound 8 and 3-hydroxypicolinic
Scheme 2. Glycosylation products derived from Ac-Gly-Lys-Gly-Phe-Leu (6) and
-glucuronic acid under dry-heating conditions (50 °C, 75% relative humidity).
acid residue for compound 9.
D
c
Assignment of signals could be interchangeable.
tions.20,21 An interesting case was the formation of a conjugated
enol–keto-immonium derivative in the reaction of sodium glucu-
ronate with the lysine-containing tripeptide, Ac-Tyr-Lys-Gly-
NH2.22
100
A possible mechanism of formation of the two novel structures
8 and 9 is shown in Scheme 3. The reaction cascade starts with the
condensation of the glucuronic acid in its open-chain form with the
free amino group of the peptide moiety to give Schiff base, which
then undergoes rearrangement to the keto-sugar peptide deriva-
tive or Amadori product 7. It can be assumed that the extra COOH
group of the sugar residue effectively catalyzes the rate-determin-
ing step in the rearrangement reaction by intramolecular proton-
ation of the imine nitrogen atom in intermediate Schiff base
leading to the immonium ion required for Amadori compound for-
mation. Water initiates and/or participates in further keto–enol
tautomerizations and carbonyl group migrations23,18 along the car-
6
7
8
9
80
60
40
20
0
bohydrate backbone leading to b- and
a-keto acid intermediates.
Unstable b-keto acid intermediate eliminates CO2 at C-6 resulting
in a 4-ketopentose. Carbonyl group migration to C-5, followed by
an intramolecular amino–carbonyl reaction and a series of dehy-
drations, yield the final 3-hydroxypyridinium compound 8. On
0
1
2
3
4
5
Time (days)
the other hand, the
lecular reaction and dehydrations, to the picolinic acid derivative 9.
a-keto acid intermediate leads, after intramo-
Figure 1. Kinetics of the formation of pyridinium compounds 8 and 9 during dry-
heating in the dark, at 50 °C, in the environment of 75% relative humidity.

ˇ
´
382
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
CHO
OH
3. Experimental
HO
3.1. General methods
OH
OH
Dry-heating experiments were performed in a CLIMACELL 111
COOH
GlcA
Cooling incubator with controlled humidity. Melting points were
determined on a Tottoli (Büchi) apparatus and are uncorrected.
Optical rotations were measured at 25 °C using an Optical Activity
LTD automatic AA-10 polarimeter. NMR spectra were recorded on
a Bruker AV 600 spectrometer, operating at 150.91 MHz for 13C and
600.13 MHZ for 1H nuclei. The spectra were measured in DMSO-d6
solution at 25 °C. Spectra were assigned based on 2D homonuclear
(COSY, NOESY and ROESY) and heteronuclear (HMQC and HMBC)
experiments. Reversed-phase high-performance liquid chromatog-
raphy (RP-HPLC) was performed on a Varian Pro Star 230 HPLC sys-
tem using a Eurospher 100 reversed-phase C-18 semipreparative
R-NH2 (6)
H2C NH
O
R
HO
OH
OH
COOH
(250 ꢁ 8 mm ID, 5
l
m) (flow rate: 1.0 ml/min) or analytical
Amadori compound 7
(150 ꢁ 4.5 mm ID, 5
lm) (flow rate: 0.5 mL/min) column under
isocratic or gradient conditions using different concentrations of
MeOH in 0.1% aq trifluoroacetic acid (TFA). UV detection was per-
formed at 215 and 254 nm using a Varian Pro Star 335 photodiode-
array detector. Mass spectrometry measurements were performed
on a Agilent Technologies 1200 HPLC system consisting of a binary
pump, a degasser, an autosampler and a diode-array detector and
coupled with a 6410 Triple Quadrupole mass spectrometer operat-
ing in a positive-ion electrospray ionization (ESI) mode. Lys-
Gly-Gly, Gly-Lys-Gly and Phe-Leu were purchased from Bachem.
H2C NH
R
H2C NH R
OH
OH
HO
HO
O
OH
COOH
OH
O
COOH
α-keto acid
β-keto acid
D-Glucuronic acid (GlcA) was purchased from Sigma Chemical Co.
3.2. Synthesis of pentapeptides
H2C NH
OH
R
OH
HO
OH
The N-terminally protected pentapeptides 1 and 4 were pre-
pared by classical solution-phase methods of peptide synthesis
with Boc-Lys(Boc)-Gly-Gly or Boc-Gly-Lys(Boc)-Gly and Phe-Leu
as the starting compounds. The coupling of the peptide segments
was achieved by using the mixed anhydride method with isobutyl
chloroformate. Boc protecting groups were cleaved by trifluoroace-
tic acid (TFA), and the crude peptides were purified by RP-HPLC.
HO
N
R
OH
COOH
O
H
OH
OH
HO
OH
3.2.1. Lys-Gly-Gly-Phe-Leu (1)
N
R
COOH
This compound was prepared from Boc-Lys(Boc)-Gly-Gly
(197 mg, 0.43 mmol) and TFA ꢁ Phe-Leu (200 mg, 0.51 mmol).
Purification by semipreparative RP-HPLC using 40% MeOH/0.1%
TFA as the eluent afforded title compound 1. Yield: 112 mg, 50%;
N
R
picolinic acid derivative 9
white solid, mp 120–130 °C; [
RP-HPLC: 37% MeOH/0.1% TFA tR 17.3 min. 1H NMR: d Lys1: 3.59
-CH), 1.58/1.67 (b-CH2), 1.37 ( -CH2), 1.52 (d-CH2), 2.75 (
CH2); Gly2: 3.78 ( -CH2), 8.78 (NH); Gly3: 3.59/3.72 (
-CH2), 8.17
(NH); Phe4: 4.42 (
-CH), 2.75/3.07 (b-CH2), 7.25 (d-CH), 7.25
-CH), 7.18 (f-CH), 8.29 (NH); Leu5: 4.06 (
-CH), 1.50 (b-CH2),
1.61 (
-CH), 0.85/0.88 (d,d0-CH3), 7.82 (NH). 13C NMR: Lys1: 52.85
-CH), 31.75 (b-CH2), 21.35 -CH2), 26.37 (d-CH2), 38.34
-CH2), 171.60 (CO); Gly2: 42.01 ( -CH2), 168.73 (CO); Gly3:
-CH2), 168.49 (CO); Phe4: 54.37 (
-CH), 37.26 (b-CH2),
-C), 129.03 (d-CH), 127.98 ( -CH), 126.15 (f-CH), 170.30
a
]
ꢂ7 (c 1.0, MeOH). Analytical
D
OH
(
a
c
e-
a
a
N
R
a
(e
a
3-hydroxypyridinium
compound 8
c
(a
(c
Scheme 3. Proposed mechanism of formation of 3-hydroxypyridinium compound
8 and picolinic acid derivative 9.
(e
a
42.01 (
137.97 (
a
c
a
e
(CO); Leu5: 51.53 (
a-CH), 41.11 (b-CH2), 24.34 (c-CH), 21.86/
The formation of ring-substituted pyridinium compounds as a
novel class of Maillard reaction products is expected to offer new
insights of biomedical importance for disorders of glucuronidation
processes as a major detoxification route for endobiotic and xeno-
biotic substances in humans. Increasing experimental evidence
indicates that 3-hydroxypyridinium epitopes are also formed dur-
ing tissue glycosylation under physiological conditions.20,24–26
Importantly, N-alkyl 3-hydroxypyridinium derivatives were identi-
fied as the minimum phototoxic chromophores contained in
endogenous hydroxypyridine photosensitizers capable of inducing
macromolecular damage.27
22.92 (d,d0-CH3), 174.64 (COOH). ESIMS m/z: [M+H]+ calcd for
C25H41N6O6: 521.31; found: 521.30.
3.2.2. Gly-Lys-Gly-Phe-Leu (4)
This compound was prepared from Boc-Gly-Lys(Boc)-Gly
(150 mg, 0.33 mmol) and TFA ꢁ Phe-Leu (150 mg, 0.36 mmol).
Purification by semipreparative RP-HPLC using 40% MeOH/0.1%
TFA as the eluent afforded title compound 4. Yield: 137 mg, 80%;
white solid, mp 165–175 °C; [
a
]
D
ꢂ35 (c 1.0, MeOH). Analytical

ˇ
´
383
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
RP-HPLC: 37% MeOH/0.1% TFA tR 15.8 min. 1H NMR: d Gly1: 3.43
-CH2); Lys2: 4.31 (
-CH), 1.49/1.69 (b-CH2), 1.30 ( -CH2), 1.52
(d-CH2), 2.73 ( -CH2), 8.21
-CH2), 8.44 (NH); Gly3: 3.63/3.72 (
(NH); Phe4: 4.39 (
-CH), 2.76/3.07 (b-CH2), 7.24 (d-CH), 7.24
-CH), 7.18 (f-CH), 8.25 (NH); Leu5: 4.02 (
-CH), 1.44/1.54
(b-CH2), 1.62 (
-CH), 0.86/0.88 (d,d0-CH3), 7.74 (NH). 13C NMR:
Gly1: 41.53 ( -CH2), 168.25 (CO); Lys2: 51.97 (
-CH), 31.29
(b-CH2), 21.89 ( -CH2), 26.46 (d-CH2), 38.45 ( -CH2), 171.37 (CO);
Gly3: 42.09 ( -CH2), 168.63 (CO); Phe4: 54.70 (
-CH), 37.13 (b-
CH2), 137.95 ( -C), 129.07 (d-CH), 128.07 ( -CH), 126.22 (f-CH),
lyophilized to yield an inseparable mixture of mono-glycated
e
(
a
a
c
a
N -(1-deoxy-
glycyl- -phenylalanyl-
deoxy- ,b- -fructofuranos-1-yluronic acid)-
phenylalanyl- -leucine (3). Analytical RP-HPLC: 37% MeOH/0.1%
TFA 2: tR 16.97 min; 3: tR 17.58 min. 1H NMR (2+3): d 1-deoxy-
,b- -fructofuranos-1-yluronic acid: 3.08/3.23, 3.03/3.12 (1-CH2),
3.74, 3.87, 3.92, 3.96, 4.09, 4.24, 4.28, 4.30, 4.36, 4.41 (3-, 4-, 5-
CH); Lys1: 3.65, 3.87, 3.96 (
-CH), 1.71, 1.82 (b-CH2), 1.34 (
CH2), 1.62 (d-CH2), 2.88, 2.92 ( -CH2),
-CH2); Gly2: 3.74/3.82 (
8.73, 8.78, 8.81 (NH); Gly3: 3.63/3.78 ( -CH2), 8.18 (NH); Phe4:
4.58 ( -CH), 2.74/3.03 (b-CH2), 7.25 (d-CH), 7.17, 7.25 ( -CH),
7.20, 7.25 (f-CH), 8.05, 8.12 (NH); Leu5: 4.23 (
-CH), 1.54 (b-CH2),
1.62 (
-CH), 0.85/0.91 (d,d0-CH3), 8.34, 8.37 (NH). 13C NMR (2+3):
1-deoxy- ,b- -fructofuranos-1-yluronic acid: 50.27, 50.44, 51.60
a,b-
D
-fructofuranos-1-yluronic acid)-
-leucine (2) and di-glycated N ,N -di-(1-
-lysyl-glycyl-glycyl-
L-lysyl-glycyl-
a
e
e
L
L
a
a
D
L
L-
(e
a
L
c
a
c
a
a
D
e
a
a
a
c-
c
e
e
a
170.22 (CO); Leu5: 51.71 (
a
-CH), 41.53 (b-CH2), 24.41 (
c-CH),
a
21.93/23.08 (d,d0-CH3), 174.87 (COOH). ESIMS m/z: [M+H]+ calcd
a
e
for C25H41N6O6: 521.31; found: 521.52.
a
c
3.2.3. Ac-Gly-Lys-Gly-Phe-Leu (6)
a
D
This peptide was synthesized manually from its C- to N-termi-
nal end by the solid-phase Fmoc method on a commercially avail-
able preloaded Fmoc-Leu Wang resin (Bachem, p-alkoxybenzyl
alcohol resin, 200–400 mesh, 0.8 mmol/g loading) on a 0.1 mmol
scale. The consecutive steps in the solid-phase peptide synthesis
performed in each cycle were (i) deprotection of the Fmoc group
by two treatments (1 and 30 min) with 20% piperidine in DMF
(v/v); (ii) coupling by applying HBTU/HOBt/DIPEA activation and
(1-CH2), 101.18, 104.20, 104.44 (2-C), 70.63, 73.11, 75,48, 76.95,
77.11, 77.40, 79.23, 79.27, 79.40, 79.72, 79.84, 81.34, 81.41, 82.11,
82.22 (3-, 4-, 5-CH), 171.84, 171.87 (6-COOH); Lys1: 51.97, 59.84,
60.13 (
24.58 (d-CH2), 47.38, 47.49, 47.57 (
Gly2: 41.79 ( -CH2), 168.03 (CO); Gly3: 41.57 (
(CO); Phe4: 53.46 (
-CH), 37.61 (b-CH2), 137.61 (
CH), 127.89 (
-CH), 126.15 (f-CH), 170.96, 171.02 (CO); Leu5:
51.42 ( -CH), 39.47 (b-CH2), 24.34 (
a
-CH), 28.94, 30.39 (b-CH2), 20.98, 21.15 (
-CH2), 167.56, 168.70 (CO);
-CH2), 168.15
-C), 129.13 (d-
c-CH2), 24.48,
e
a
a
a
c
e
a
threefold excess of the appropriate Fmoc-amino acid for
a
c
-CH), 20.57/22.24 (d,d0-CH3),
30 min, and in the last step, by applying the Ac-Gly for 1 h and
(iii) removal of the peptide from the resin by treatment with a mix-
ture of TFA:TIS:H2O in a ratio of 9.5:0.25:0.25 (v/v/v) for 1 h. Suc-
cessive deprotection and coupling steps were monitored by the
positive and negative Kaiser (ninhydrin) test, respectively. The
peptide 6 was obtained as a filtrate in TFA, and it was precipitated
with cold dry diisopropyl ether. Purification by semipreparative
RP-HPLC using 40% MeOH/0.1% TFA as the eluent afforded title
173.68, 174.01 (COOH). ESIMS m/z: [M+H]+ calcd for C31H49N6O12
697.34; found: 697.10; m/z: [M+2H]+ calcd for C37H58N6O18
874.89; found: 874.26.
:
:
3.3.2. N-(1-Deoxy-
a
,b-
-fructofuranos-1-yluronic acid)-
-leucine (5)
Compound 5 was obtained starting from the diacetate salt of
Gly-Lys-Gly-Phe-Leu (4) (32 mg, 0.05 mmol) and -glucuronic acid
D
-fructofuranos-1-yluronic acid)-glycyl-
e
N -(1-deoxy-a,b-
D
L-lysyl-glycyl-
L-phenylalanyl-
L
compound 6. Yield: 31 mg, 56%; [
RP-HPLC: 37% MeOH/0.1% TFA tR 27.78 min. 1H NMR: d Ac-Gly1:
1.84 (Ac, CH3), 3.70 ( -CH), 1.52/
-CH2), 8.16 (NH); Lys2: 4.23 (
1.64 (b-CH2), 1.29 ( -CH2), 1.52 (d-CH2), 2.73 ( -CH2), 8.07 (NH);
Gly3: 3.55/3.72 ( -CH2), 8.17 (NH); Phe4: 4.49 (
-CH), 2.77/3.05
(b-CH2), 7.24 (d-CH), 7.24 (
-CH), 7.18 (f-CH), 8.04 (NH); Leu5:
4.13 ( -CH), 1.52 (b-CH2), 1.64 (
-CH), 0.85/0.89 (d,d0-CH3), 8.03
(NH). 13C NMR: Ac-Gly1: 22.41 (Ac, CH3), 42.01 (
-CH2), 169.16
(CO), 169.75 (Ac, CO); Lys2: 52.21 (
-CH), 31.14 (b-CH2), 21.98
-CH2), 171.75 (CO); Gly3: 42.11
-CH2), 168.45 (CO); Phe4: 54.00 (
-CH), 37.48 (b-CH2), 137.81
-C), 129.20 (d-CH), 128.03 ( -CH), 126.22 (f-CH), 170.60 (CO);
a]
D ꢂ21 (c 1.0, MeOH). Analytical
D
(145 mg, 0.75 mmol) by using the same procedure as described for
compounds 2 and 3. Purification by semipreparative RP-HPLC
using 36% MeOH/0.1% TFA as the eluent afforded the pure title
a
a
c
e
a
a
compound 5. Yield: 17 mg, 40%; white hygroscopic solid; [
a
]
ꢂ3
D
e
(c 1.0, MeOH). Analytical RP-HPLC: 33% MeOH/0.1% TFA tR
25.0 min. 1H and 13C NMR data are given in Table 1. ESIMS m/z:
[MꢂH2O+H]+ calcd for C37H55N6O17: 855.36; found: 855.24.
a
c
a
a
e
(
(
(
c-CH2), 26.43 (d-CH2), 38.53 (e
a
3.3.3. N-Acetyl-glycyl-N -(1-deoxy-
a
,b-
-phenylalanyl-
Compound 7 was obtained from the acetate salt of Ac-Gly-Lys-
Gly-Phe-Leu (6) (31 mg, 0.05 mmol) and -glucuronic acid
D
-fructofuranos-1-yl-
a
uronic acid)-
L
-lysyl-glycyl-
L
L
-leucine (7)
c
e
Leu5: 51.13 (
a
-CH), 40.66 (b-CH2), 24.35 (
c
-CH), 21.66/22.97
D
(d,d0-CH3), 174.41 (COOH). ESIMS m/z: [M+H]+ calcd for
(145 mg, 0.75 mmol) by using the same procedure as described
for compound 5. Purification by semipreparative RP-HPLC using
40% MeOH/0.1% TFA as the eluent afforded the pure title com-
C27H43N6O7: 563.32; found: 563.40.
3.3. Synthesis of
compounds
D-glucuronic acid-derived Amadori
pound 7. Yield: 29 mg, 79%; white hygroscopic solid; [
a
]
ꢂ2 (c
D
1.0, MeOH). Analytical RP-HPLC: 37% MeOH/0.1% TFA tR 25.1 min.
1H and 13C NMR data are given in Table 1. ESIMS m/z: [M+H]+ calcd
for C33H51N6O13: 739.35; found: 739.12.
e
3.3.1. N -(1-Deoxy-
a
,b-
-phenylalanyl-
-fructofuranos-1-yluronic acid)-
-phenylalanyl- -leucine (3)
-Glucuronic acid (145 mg, 0.75 mmol) and Lys-Gly-Gly-Phe-
D
-fructofuranos-1-yluronic acid)-
L
-lysyl-
-leucine (2) and N ,N -di-(1-
-lysyl-glycyl-
a
e
glycyl-glycyl-
deoxy- ,b-
glycyl-
L
L
3.4. Synthesis of
D-glucuronic acid-derived pyridinium
a
D
L
compounds 8 and 9
L
L
D
D-Glucuronic acid (582 mg, 3.0 mmol) and Ac-Gly-Lys-Gly-Phe-
Leu (1), as diacetate salt, (32 mg, 0.05 mmol) were dissolved in
dry MeOH (30 mL), and the reaction mixture was kept in a closed
round-bottomed flask for three days at 50 °C. The solvent was
evaporated, and the excess of sugar from the reaction mixture
was removed by using an C-18 solid-phase extraction (SPE)
cartridge. The cartridge was first eluted with water to remove the
sugar. The peptide material was then recovered with MeOH.
The effluent was evaporated, and the residue obtained was purified
by semipreparative RP-HPLC using 33% MeOH/0.1% TFA and
Leu (6), as acetate salt (124 mg, 0.2 mmol), were dissolved in water
(3 mL) and were lyophilized to give a friable glassy material with a
water content of 9%. The reactants were then exposed to 75% rela-
tive humidity at 50 °C, in the dark, for five days. The dark brown
solid mixture was dissolved in water and was applied to an SPE ca-
tridge to remove the excess sugar. Purification of the peptide-con-
taining fraction, recovered by MeOH, by semipreparative RP-HPLC
using 37% MeOH/0.1% TFA as the eluent afforded Amadori

ˇ
´
Štefica Horvat, M. Rošcic / Carbohydrate Research 345 (2010) 377–384
384
compound 7 (8 mg, 5%), 3-hydroxypyridinium derivative 8 (12 mg,
9%) and 3-hydroxypicolinic acid derivative 9 (18 mg, 13%).
References
1. Zhang, Q. B.; Ames, J. M.; Smith, R. D.; Baynes, J. W.; Metz, T. O. J. Proteome Res.
2009, 8, 754–769.
2. Price, C. L.; Knight, S. C. Curr. Pharm. Design 2007, 13, 3681–3687.
3. Brownlee, M. Annu. Rev. Med. 1995, 46, 223–234.
4. Schalkwijk, C. G.; Stehouwer, C. D. A.; van Hinsbergh, V. W. M. Diabetes Metab.
Res. Rev. 2004, 20, 369–382.
5. Frolov, A.; Hoffmann, P.; Hoffmann, R. J. Mass Spectrom. 2006, 41, 1459–1469.
3.4.1. 3-Hydroxypyridinium derivative (8)
White hygroscopic solid; [
a
]
D ꢂ13 (c 0.9, MeOH). Analytical RP–
HPLC: 37% MeOH/0.1% TFA tR 34.5 min. 1H and 13C NMR data are
given in Table 2. ESIMS m/z: [M+H]+ calcd for C32H45N6O8:
641.33; found: 641.08.
ˇ ´
6. Rošcic, M.; Horvat, Š. Bioorg. Med. Chem. 2006, 14, 4933–4943.
7. Glucuronidation of Drugs and Other Compounds; Dutton, G. J., Ed.; CRC Press:
Boca Raton, FL, 1980.
8. Bailey, M. J.; Dickinson, R. G. Chem. Biol. Interact. 2003, 145, 117–137.
9. Spahn-Langguth, H.; Benet, L. Z. Drug Metab. Rev. 1992, 24, 5–47.
10. Faed, E. M. Drug Metab. Rev. 1984, 15, 1213–1249.
11. Akira, K.; Uchijima, T.; Hashimoto, T. Chem. Res. Toxicol. 2002, 15, 765–772.
12. Baba, A.; Yoshioka, T. Chem. Res. Toxicol. 2009, 22, 158–172.
13. Asif, A. R.; Armstrong, V. W.; Voland, A.; Wieland, E.; Oellerich, M.; Shipkova,
M. Biochimie 2007, 89, 393–402.
3.4.2. 3-Hydroxypicolinic acid derivative (9)
White hygroscopic solid; [
a
]
D ꢂ12 (c 1.0, MeOH). Analytical RP–
HPLC: 37% MeOH/0.1% TFA tR 32.7 min. 1H and 13C NMR data are
given in Table 2. ESIMS m/z: [M+H]+ calcd for C33H45N6O10
:
685.32; found: 685.07.
14. Stachulski, A. V.; Harding, J. R.; Lindon, J. C.; Maggs, J. L.; Park, B. K.; Wilson, I. D.
J. Med. Chem. 2006, 49, 6931–6945.
15. Wang, J.; Davis, M.; Li, F.; Azam, F.; Scatina, J.; Talaat, R. Chem. Res. Toxicol.
2004, 17, 1206–1216.
3.5. Kinetics of pyridinium compounds formation under dry-
heating conditions
16. Smith, P. C.; Wang, C. Biochem. Pharmacol. 1992, 44, 1661–1668.
17. Chiou, Y.; Tomer, K. B.; Smith, P. C. Chem. Biol. Interact. 1999, 121, 141–159.
D
-Glucuronic acid (14.5 mg, 0.075 mmol) and Ac-Gly-Lys-Gly-
Phe-Leu (6), as acetate salt (3 mg, 0.005 mmol), were dissolved in
water (0.25 mL) and were lyophilized. The peptide–carbohydrate
sample was then exposed to 75% relative humidity at 50 °C, in
the dark, for various periods up to five days. The concentrations
of the starting peptide and glycation products in the incubation
mixture were determined by RP-HPLC on an analytical column at
a flow rate of 0.5 mL/min with 37% MeOH/0.1% TFA by using an
external standard method. UV detection was performed at 215
and 254 nm.
ˇ ´
18. Horvat, Š.; Rošcic, M.; Lemieux, C.; Nguyen, T. M.-D.; Schiller, P. W. Chem. Biol.
Drug Des. 2007, 70, 30–39.
19. Miao, S.; Roos, Y. H. Innov. Food Sci. Emerg. Technol. 2006, 7, 182–194.
20. Tessier, F. J.; Monnier, V. M.; Sayre, L. M.; Kornfield, J. A. Biochem. J. 2003, 369,
705–719.
21. Reihl, O.; Biemel, K. M.; Lederer, M. O.; Schwack, W. Carbohydr. Res. 2004, 339,
705–714.
22. Takeda, Y.; Kyogoku, Y.; Ishidate, M. Carbohydr. Res. 1977, 59, 363–377.
23. Reihl, O.; Rothenbacher, T. M.; Lederer, M. O.; Schwack, W. Carbohydr. Res.
2004, 339, 1609–1618.
24. Nagai, R.; Hayashi, C. M.; Xia, L.; Takeya, M.; Horiuchi, S. J. Biol. Chem. 2002, 277,
48905–48912.
25. Liu, Z. F.; Sayre, L. M. Chem. Res. Toxicol. 2003, 16, 232–241.
26. Argirov, O. K.; Lin, B.; Ortwerth, B. J. J. Biol. Chem. 2004, 279, 6487–6495.
27. Wondrak, G. T.; Roberts, M. J.; Jacobson, M. K.; Jacobson, E. L. J. Biol. Chem. 2004,
279, 30009–30020.
Acknowledgements
This work was supported by the Croatian Ministry of Science,
Education and Sports, Grant No. 098-0982933-2936. The authors
ˇ ´
thank Mrs. Milica Perc and Miss Danijela Mikulcic for excellent
technical assistance.