Glutathione reacts with glyoxal at the N-terminal

Article abstract of DOI:10.1271/bbb.90340

The elevation of such dicarbonyl compounds as glyoxal and the depletion of GSH occur simultaneously in diabetic patients. Enabling a nonenzymatic glycation reaction with GSH and glyoxal is therefore proposed. However, the reaction mechanism for GSH and glyoxal has not been precisely defined. We isolated in this study the major products obtained by the reaction of GSH and glyoxal under physiological conditions, and clarified the chemical structure of these compounds by MS and NMR analyses for the first time. We identified the major product after 24 h as N-[3-(2, 5-dioxomorpholin-3-yl)- propanoyl] cysteinylglycine, and the one after 30 min as N-glycoloyl-γ- glutamylcysteinylglycine (the intermediate of the former compound). Our results suggest that GSH reacted with glyoxal at the α-NH₂ group of the glutamate residue, but not at the SH group of the cysteine residue.

Full text of DOI:10.1271/bbb.90340

Biosci. Biotechnol. Biochem., 73 (11), 2408–2411, 2009
Glutathione Reacts with Glyoxal at the N-Terminal
y
1
Yuri NOMI,1 Haruko AIZAWA,1 Tadao KURATA,2 Kazutoshi SHINDO,1; and Chuyen Van NGUYEN
1Department of Food and Nutrition, Japan Women’s University, 2-8-1 Mejirodai, Bunkyo-ku, Tokyo 112-8681, Japan
2Department of Applied Sciences, Niigata University of Pharmacy and Applied Sciences,
265-1 Higashijima, Akiba-ku, Niigata 956-8603, Japan
Received May 12, 2009; Accepted July 17, 2009; Online Publication, November 7, 2009
[doi:10.1271/bbb.90340]
The elevation of such dicarbonyl compounds as
glyoxal and the depletion of GSH occur simultaneously
in diabetic patients. Enabling a nonenzymatic glycation
reaction with GSH and glyoxal is therefore proposed.
However, the reaction mechanism for GSH and glyoxal
has not been precisely defined. We isolated in this study
the major products obtained by the reaction of GSH and
glyoxal under physiological conditions, and clarified the
chemical structure of these compounds by MS and
NMR analyses for the first time. We identified the major
product after 24 h as N-[3-(2,5-dioxomorpholin-3-yl)-
propanoyl]cysteinylglycine, and the one after 30 min as
N-glycoloyl-ꢀ-glutamylcysteinylglycine (the intermedi-
ate of the former compound). Our results suggest that
GSH reacted with glyoxal at the ꢁ-NH2 group of the
glutamate residue, but not at the SH group of the
cysteine residue.
that these Amadori compounds inhibited glutathione
peroxidase.9) Linetsky et al. have shown that the
reaction of GSH with glucose (or fructose) at the
ꢀ-NH2 group of the glutamate residue in GSH resulted
in the formation of Amadori (or Heynes) compounds,
and that the products inhibited several glutathione-
related enzymes.10,11) Since these enzymes contribute
to the cell defence system against toxic reactive
species,12,13) glycation of GSH may lead to diabetic
complications.
Since the elevation of glyoxal and depletion of
GSH in the plasma of diabetic patients have been
reported,7,8,14) we hypothesized that glyoxal may be a
reactive molecule for GSH and investigated the reaction
between glyoxal and GSH. In this present study, we
isolated the major products produced by the reaction
with GSH and glyoxal under physiological conditions by
preparative HPLC, and determined their structures by
MS and NMR. A novel and specific reaction mechanism
between GSH and glyoxal is reported here.
Key words: glutathione; glyoxal; glycation; morpholine
GSH is the major low-molecular-mass thiol com-
pound in plants and animals. It is typically present in
high concentrations (0.1–10 mM) in the cytosol,1) pro-
tects cells from the toxic effects of reactive oxygen
species and maintains a cellular redox status.2) In
addition, GSH is also involved in the metabolism of
foreign objects.
Materials and Methods
Materials. GSH was purchased from Wako Pure Chemical
Industries (Osaka, Japan). A 40% aqueous glyoxal solution and the
other reagents were purchased from Kanto Kagaku (Tokyo, Japan).
Incubation of GSH with glyoxal. GSH (10 mM) was incubated with
glyoxal (10 mM) in a 100 mM sodium phosphate buffer (pH 7.4) at
37 ꢀC for up to 24 h. The progress of the reaction was monitored by an
HPLC analysis of the diluted solution (10 times by 20 mM HCl) under
the following conditions: column, Inertsil Peptides C18 (4.6 mm
i.d. ꢁ 250 mm, GL Sciences, Japan); solvent, 0.1% trifluoroacetic
acid in water (solvent A); 0.1% trifluoroacetic acid in acetonitrile
(solvent B); linear gradient of B, 2–20% in 30 min; flow rate,
1.0 ml/min; detection, 200–370 nm (diode array); temperature, 40 ꢀC.
HPLC was monitored at reaction times of 5 min, 30 min, 1 h, 2 h, 6 h,
and 24 h.
A chemical reaction between carbohydrates and
proteins (nonenzymatic glycation) frequently occurs in
the plasma. This reaction is accelerated in the state of
hyperglycemia in diabetes, and advanced glycation
end-products (AGEs) are produced through this reaction.
AGEs have been reported to be correlated with the
progression of diabetes and aging.3) In addition to
carbohydrates, highly reactive ꢀ-dicarbonyl compounds
have also been shown to react with proteins and initiate
the formation of AGEs.4) Among the ꢀ-dicarbonyl
compounds, glyoxal, which is formed by the degradation
of glycated proteins, oxidative degradation of glucose,
lipid peroxidation, and ascorbate autoxidation,5,6) has
been proposed to be one of the main reactants for AGEs.
Elevation of the glyoxal concentration in the plasma of
diabetic and uremic patients has also been reported.7,8)
A reaction concerning the nonenzymatic glycation of
GSH has recently been observed. Januel et al. have
reported that ribose formed Amadori compounds with
the ꢀ-NH2 group of the glutamate residue in GSH, and
Isolation of the GSH-glyoxal reaction product. 307 mg (1 mmol) of
GSH in 100 ml of a sodium phosphate buffer (100 mM, pH 7.4) was
incubated with glyoxal (1 mmol) at 37 ꢀC for 30 min or 24 h. After
filtration being passed through a 0.20-mm filter (Advantec, Japan), the
reaction mixture was concentrated to dryness under reduced pressure.
The concentrate was extracted with 10 ml of a dichloromethane-
methanol-water (3:1:0.2) solution, and the supernatant was evaporated
to dryness. This extract was dissolved in 2 ml of 50 mM hydrochloric
acid, and each 40 ml of the solution was purified by preparative HPLC
under the following conditions: column, Inertsil ODS-3 (10 mm
i.d. ꢁ 250 mm, GL Sciences, Japan); solvent, 0.1% trifluoroacetic acid
y
To whom correspondence should be addressed. Tel/Fax: +81-3-5981-3433; E-mail: kshindo@fc.jwu.ac.jp
Abbreviations: AGE, advanced glycation end-product; HRESI-MS, high resolution electrospray ionization mass spectrometry

GSH Reacts with Glyoxal at the N-Terminal
in water (solvent A); 0.1% trifluoroacetic acid in acetonitrile
(solvent B); linear gradient of 2–15% in 30 min; flow rate,
2409
B
5 ml/min; detection, 230 nm; temperature, 40 ꢀC. The peaks at
retention times of 9.6 min (30-min reaction solution) and 14.5 min
(24-h reaction solution) were collected and lyophilized to give 12.8 mg
(compound 1, 5.8% yield) and 50.6 mg (compound 2, 25% yield) of
pure materials.
Structural analysis. High-resolution electrospray ionization mass
spectrometry (HRESI-MS) was recorded with a Jeol JMS-T100LP.
1H- and 13C-NMR spectra were measured with a Bruker AMX400.
Chemical shifts are given in ppm, using DSS as an external standard.
Physico-chemical properties. N-[3-(2,5-dioxomorpholin-3-yl)pro-
panoyl]cysteinylglycine (compound 2) HRESI-MS m=z ðM þ NaÞþ
:
calcd. for C12H17N3O7S, 370.06849; found, 370.06896. The 1H- and
13C-NMR data for compound 2 are summarized in Table 1. N-
glycoloyl-ꢁ-glutamylcysteinylglycine (compound 1) HRESI-MS m=z
ðM ꢂ HÞꢂ: calcd. for C12H19N3O8S, 364.08146; found, 364.08135. The
1H-NMR data for compound 1 are summarized in Table 1.
Results and Discussion
To clarify the reaction mechanism for GSH with
glyoxal, the reaction products of GSH with or without
glyoxal under physiological conditions (pH 7.4, 37 ꢀC,
up to 24 h) were investigated by a diode array detection-
HPLC analysis. The results are shown in Fig. 1. In a
control sample (without glyoxal), approximately 30% of
GSH changed to GSSG within 24 h, while a rapid time-
dependent decrease in GSH and the formation of new
reaction products were observed in an experimental
sample. After 24 h of incubation, about 60% of GSH was
converted to reaction products. In the HPLC analysis,
two main peaks (retention times of 11.4 min and
14.8 min) were observed. The peak at retention time
11.4 min (compound 1) appeared after 5 min, reached a
maximum after 30 min, and disappeared after 2 h, while
the peak at retention time 14.8 min (compound 2)
appeared after 1 h and reached a maximum after 24 h
(Fig. 1). These observations indicated that these two
compounds could be related to each other. To confirm
this, each compound was purified by the procedure
described in the Materials and Methods section, en-
abling 12.8 mg of pure 1 and 50.6 mg of pure 2 to be
obtained. The structural analysis was started with 2
because it was more stable than 1 in a D2O solution.
The molecular formula for 2 was determined as
C12H17N3O7S (GSH þ C2O) by HRESI-MS, corre-
sponding to the addition of one molecule of glyoxal to
GSH and the loss of one molecule of water.
Fig. 1. Chromatograms of the Incubation Mixture of GSH with
Glyoxal.
A, HPLC profiles obtained by the incubation of GSH (10 mM) and
glyoxal (10 mM). A-1, 5 min; A-2, 30 min; A-3, 1 h; A-4, 2 h; A-5,
6 h; A-6, 24 h. B, HPLC profile obtained by the incubation of GSH
(10 mM) without glyoxal. B-1, 0 min; B-2, 24 h.
respectively, by the observation of deuterium-induced
upfield shifts at C-400 and C-700 in the 13C-NMR
spectrum of 2 taken in D2O/H2O (no upfield shifts
were observed at C-500 or C-600).15) All of these
observations allowed the structure of 2 to be deter-
mined as shown in Fig. 2. Thus, compound 2 was
identified as N-[3-(2,5-dioxomorpholin-3-yl)propanoyl]-
cysteinylglycine and was considered to be a novel
1
compound according to the CAS database. The H- and
13C-NMR data for compound 2 are summarized in
Table 1.
The molecular formula of 1 was determined to be
C12H19N3O8S (2 þ H2O) by HRESI-MS data. Since 1
was unstable in D2O, the structure of 1 was analysed by
a comparison of the 1H-NMR spectrum of 1 with that of
1
1
1
The H-NMR, 13C-NMR, H-1H COSY, and HMQC
2. In the comparison, all the H signals derived from
spectra of 2 in D2O were measured and compared with
1
glutamate, cysteine, and glycine in 2 were conserved in
1. Signals of an isolated methylene (H-600) of 1 were
observed as a singlet and in a higher field (ꢂ 3.96)
compared to the corresponding signals of 2 (ꢂ 3.94 (d,
J ¼ 16:5 Hz) and ꢂ 4.28 (d, J ¼ 16:5 Hz)). These
findings enabled the structure of 1 to be determined
as N-glycoloyl-ꢁ-glutamylcysteinylglycine which is a
novel compound according to the CAS database. The
1H-NMR data for compound 1 are summarized in
Table 1.
those of GSH. The comparison proved that all the H
and 13C signals in GSH (derived from glutamate,
cysteine, and glycine) were present in 2. Signals of one
carbonyl carbon (ꢂ 172.9, C-700) and one isolated
methylene (ꢂ 48.3, C-600) were observed only in 2. An
1
HMBC experiment on 2 proved the H-13C long-range
coupling from H-600 (ꢂ 3.94 and ꢂ 4.28, isolated
methylene) to C-500 (ꢂ 182.6) and C-700, and from H-300
(ꢂ 2.45, ꢃ-methylene of Glu) to C-500. Considering the
molecular formula of 2, the presence of
a
2,
The formation of 1 and 2 can be explained according
to the reaction scheme shown in Fig. 3. The addition of
glyoxal to the N-terminal amino group of GSH may
have led to keto-enol tautomerism, and 1 may have been
5-dioxomorpholine structure containing C-400, C-500,
C-600, and C-700 in 2 was thus proposed. C-700 and C-500
were assigned to an amide carbonyl and ester carbonyl,

2410
Y. NOMI et al.
SH
O
SH
O
O
O
O
1
H
N
H
3'
H2N
HO
H
N
H
N
3"
H
1
2
N
H
OH
O
Glyoxal
O
GSH
1'
+
N
H
OH
1"
2'
O
7"
4"
2"
2
O
COOH
5"
O
6"
OH
SH
SH
O
O
O
O
1
H
N
H
N
3'
H
H
N
H
N
3"
N
H
OH
O
1'
O
HO
O
N
H
OH
2'
COOH
COOH
O
1"
4"
7"
2
2"
H
O
O
6"
5"
SH
O
O
H
N
H
N
Fig. 2. Structures of N-Glycoloyl-ꢁ-glutamylcysteinylglycine (1) and
N-[3-(2,5-Dioxomorpholin-3-yl)propanoyl]cysteinylglycine (2).
HO
HO
N
H
OH
O
H
Table 1. 1H- and 13C-NMR Data for N-Glycoloyl-ꢁ-glutamylcystei-
nylglycine (1) and N-[3-(2,5-Dioxomorpholin-3-yl)propanoyl]cystei-
nylglycine (2) in D2O
SH
O
O
O
O
H
N
H
N
O
O
1
2
1
2
N
H
OH
OH
COOH
O
Position
ꢂH
ꢂH
ꢂC
OH
1
175.6
43.8
- H2O
2
3.96 (s, 2H)
3.98 (s, 2H)
SH
10
200
3
100
200
300
174.8
58.0
28.1
H
N
H
N
4.53 (t, 1H, J ¼ 6:0 Hz) 4.57 (t, 1H, J ¼ 5:7 Hz)
N
H
2.91(m, 2H)
2.91 (t, 2H, J ¼ 5:7 Hz)
O
178.2
31.7
25.3
O
O
2.45–2.59 (2H)
2.15 (m, 1H),
2.55 (m, 1H)
4.29 (m, 1H)
2.40–2.50 (2H)
2.23 (m, 1H),
2.45 (m, 1H)
4.44 (m, 1H)
Fig. 3. Proposed Pathway for the Formation of N-Glycoloyl-ꢁ-
glutamylcysteinylglycine (1) and N-[3-(2,5-Dioxomorpholin-3-yl)-
propanoyl]cysteinylglycine (2) by the Reaction of GSH with
Glyoxal.
400
500
600
64.0
182.6
3.96 (s, 2H)
3.94 (d, 1H, J ¼ 16:5 Hz), 48.3
4.28 (d, 1H, J ¼ 16:5 Hz)
172.9
Linetsky et al. have reported that the reaction
products of glucose (or fructose) and GSH [Amadori
(or Heynes) compounds] did not work efficiently as the
substrate of glutathione peroxidase and glutathione
S-transferase because of the loss of the NH2 group in
GSH which was utilized to form hydrogen bonds
between GSH and the enzyme.10,11) Since 1 and 2 had
also lost the NH2 group, these compounds may not work
as substrates for the foregoing enzymes.
In conclusion, this study has clarified the reaction
mechanism between GSH and glyoxal for the first time.
Before this study, we thought that the SH group of the
cysteine residue in GSH may have been the primary
reaction target for glyoxal; however, the corresponding
reaction products were not obtained, suggesting that the
reaction of N-terminal modification occurred favorably
in the reaction between GSH and glyoxal. Although the
physiological roles of 1 and 2 have not yet been tested,
the depletion of GSH in cytosol could at least contribute
to an increase in oxidative stress.
700
produced. The subsquent dehydrating condensation
between the remaining hydroxyl group and carboxyl
group of the GSH glutamyl moiety (the formation of a
lactone) may have resulted in the formation of a
dioxomorpholine derivative (2).
N-terminal modification of GSH induced by glyoxal
was clarified for the first time in this study. In a previous
study, Zeng and Davies detected a molecular ion peak
ðM þ HÞþ at m=z 348 from an ESI-MS analysis of the
incubated mixtures of N-acetyl-cysteine, GSH and
glyoxal, suggesting the formation of a reaction product
between glyoxal and GSH.16) However, the chemical
structure of this peak was not determined by them. Since
the molecular weight of 2 was 347, it is possible that
their ion peak was identical to 2.
Krause et al. have demonstrated that N-terminal
modification of the model peptide, Gly-Ala-Phe,
induced by glyoxal under physiological conditions
resulted in the formation of a pyrazinone-type structure
possessing a characteristic UV absorption maximum at
322 nm.17) Chuyen et al. also reported the same reaction
between the tripeptide, Leu-Gly-Gly, and glyoxal under
food processing conditions.18) However, such a charac-
teristic UV absorption compound was not found in the
present study. The difference in the reaction was due to
the ꢁ-glutamyl residue in GSH. Therefore, a dioxomor-
pholine-type product like compound 2 would probably
be a specific product when glyoxal reacts with GSH.
References
1) Meister A, J. Biol. Chem., 263, 17205–17208 (1988).
2) Sies H, Free Radic. Biol. Med., 27, 916–921 (1999).
3) Thorpe SR and Baynes JW, Amino Acids, 25, 275–281 (2003).
4) Miyata T, Kurokawa K, and Van Ypersele De Strihou C, J. Am.
Soc. Nephrol., 11, 1744–1752 (2000).
5) Mlakar A, Batna A, Dudda A, and Spiteller G, Free Radic. Res.,
25, 525–539 (1996).
6) Wells-Knecht KJ, Zyzak DV, Litchfield JE, Thorpe SR, and
Baynes JW, Biochemistry, 34, 3702–3709 (1995).

GSH Reacts with Glyoxal at the N-Terminal
2411
7) Odani H, Shinzato T, Matsumoto Y, Usami J, and Maeda K,
Biochem. Biophys. Res. Commun., 256, 89–93 (1999).
8) Lapolla A, Flamini R, Dalla Vedova A, Senesi A, Reitano R,
Fedele D, Basso E, Seraglia R, and Traldi P, Clin. Chem. Lab.
Med., 41, 1166–1173 (2003).
13) Muruganandam A, Tannous M, and Mutus B, Clin. Biochem.,
27, 293–298 (1994).
14) Samiec PS, Drews-Botsch C, Flagg EW, Kurtz JC, Sternberg
PJr, Reed RL, and Jones DP, Free Radic. Biol. Med., 24, 699–
704 (1998).
9) Januel C, Fay LB, Ruggiero D, Lagarde M, and Vericel E,
Biochim. Biophys. Acta, 1620, 125–132 (2003).
15) Pfeffer PE, Valentine KM, and Parrish FW, J. Am. Chem. Soc.,
101, 1265–1279 (1979).
10) Linetsky MD, Shipova EV, Legrand RD, and Argirov OO,
Biochim. Biophys. Acta, 1724, 181–193 (2005).
16) Zeng J and Davies MJ, Chem. Res. Toxicol., 19, 1668–1676
(2006).
11) Linetsky M, Shipova EV, and Argirov OK, Arch. Biochem.
Biophys., 449, 34–46 (2006).
17) Krause R, Kuhn J, Penndorf I, Knoll K, and Henle T, Amino
Acids, 27, 9–18 (2004).
12) Blakytny
(1992).
R
and Harding JJ, Biochem. J., 288, 303–307
18) Chuyen NV, Kurata T, and Fujimaki M, Agric. Biol. Chem., 37,
1613–1618 (1973).