- Email: [email protected]
Covalent coupling of reduced glutathione with ribose: loss of cosubstrate ability to glutathione peroxidase
Biochimica et Biophysica Acta 1620 (2003) 125 – 132 www.bba-direct.com
Covalent coupling of reduced glutathione with ribose: loss of cosubstrate ability to glutathione peroxidase Caroline Januel a, Laurent B. Fay b, Daniel Ruggiero c, Michel Lagarde a, Evelyne Ve´ricel a,* a
INSERM U 352, Biochimie et Pharmacologie, INSA-Lyon, Baˆt. Louis Pasteur, 69621 Villeurbanne Cedex, France b Nestec Ltd, Nestle´ Research Centre, Vers chez les Blancs, P.O. Box 44 1000 Lausanne, 26, Switzerland c MERCK-INSERM U 352, INSA-Lyon, Baˆt. Louis Pasteur, 69621 Villeurbanne, France Received 1 October 2002; received in revised form 2 December 2002; accepted 4 December 2002
Abstract Glycation (nonenzymatic glycosylation of proteins) is known to be increased as a result of hyperglycaemia in diabetes. Moreover, cell glutathione concentration has been found to be lower in diabetics and such depletion may impair the cell defence against toxic radical species. Ribose being a potent reducing sugar expected to be increased in cells of diabetics where the pentose phosphate pathway is enhanced, its putative condensation with glutathione was investigated. Reduced glutathione (GSH) was incubated with ribose and the structure of the resultant product was assessed by mass spectrometry, as well as the measurement of its remaining thiol group. A covalent reaction clearly occurred between the reducing sugar and GSH, to give an adduct named N-ribosyl-1-glutathione. This adduct appears to be the Amadori product resulting from the condensation of the primary amine group of GSH with the aldehyde group of ribose. Interestingly, the adduct could not be used as a proper substrate by glutathione peroxidase although it keeps its thiol group. We conclude that the coupling of GSH with a monosaccharide such as ribose might contribute to the decreased cell GSH and glutathione peroxidase activity observed in diabetics. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Amadori; Diabetes; Glutathione; Glycation; Mass spectrometry; Ribose
1. Introduction Mammalian cells contain high concentrations of glutathione (g-glutamyl-cysteinyl-glycine). It is the prevalent nonprotein thiol and the most abundant low molecular weight peptide (usually in the millimolar range) [1]. Glutathione is mainly found in its reduced form (GSH) and, in much smaller amounts, in its oxidized one. GSH, a major endogenous protective system, is involved in the conjugation of several types of compounds, protects cells from the toxic effects of reactive oxygen species and maintains cellular redox status
Abbreviations: cGPx, cytosolic glutathione peroxidase; EDTA, ethylenediaminetetraacetic acid; GSH, reduced glutathione; GSSG, oxidized glutathione; HPLC, high-performance liquid chromatography; MS, mass spectrometry; MS/MS, tandem mass spectrometry; NADPH, nicotinamide adenine dinucleotide phosphate (reduced form) * Corresponding author. Tel.: +33-4-72-43-84-79; fax: +33-4-72-4385-24. E-mail address: [email protected]. (E. Ve´ricel).
[2]. Finally, GSH is a cosubstrate of several enzymes including GSH peroxidase (GPx). Nonenzymatic glycosylation (glycation) results from the chemical reaction between carbohydrates and proteins. The reaction is accelerated in response to hyperglycaemia in diabetes and is involved in the pathogenesis of its chronic complications [3]. The very early products of the glycation process such as the Schiff base and Amadori products result from the covalent addition of glucose (or other reducing carbohydrates) to protein amino groups (Fig. 1). The first protein demonstrated to be glycated in vivo was haemoglobin [4], and the fraction of glycated haemoglobin increased from 5% in normal people to 10% in diabetic patients. Indeed, it is considered that the level of glycated haemoglobin also reflects the blood glucose concentration of such patients [5]. Since this discovery, glycation of a number of proteins (collagen, albumin, crystalline, etc.) [6] and several enzymes (erythrocyte superoxide dismutase, erythrocyte purine nucleoside phosphorylase, platelet GPx, etc.) [7– 9] has been reported. Such a glycation often impairs the function of proteins and contributes to inhibition of enzyme
0304-4165/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-4165(02)00525-1
126
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
Fig. 1. Early products in protein condensation with glucose.
activity [7]. Amadori products may be transformed in a variety of compounds until advanced glycosylation endproducts (AGE). The AGE formation is known to be affected by different factors, including metal ions which catalyze both the glycoxidation and the conversion of Amadori products [10]. Both early glycation Amadorimodified protein [11] and AGE [12] formation were found increased in diabetes. Moreover, in diabetes, high glucose concentrations lead to both formation of toxic reactive oxygen species [13] and alteration in cell defence systems such as GSSG-reductase [8], GPx [9] and superoxide dismutase [6]. Hence, the protection against toxic reactive species is lower in diabetics and GSH maintenance is impaired [14,15]. One working hypothesis to explain the decrease of GSH in diabetes could be the glycation of GSH. To determine whether GSH could be first condensed with a carbohydrate, it was incubated with ribose, a potent glycating agent assumed to be increased in diabetes. The condensation product was isolated by HPLC and its structure determined by mass spectrometry (MS) and named N-ribosyl-1-glutathione. Finally, N-ribosyl-1-glutathione was tested as a cosubstrate for cytosolic GPx (cGPx) instead of GSH.
2. Methods 2.1. Materials Glutathione (reduced form), D( )-ribose, NADPH, bovine erythrocyte cGPx, GSSG-reductase, tert-butyl hydro-
peroxide and other reagents were obtained from SigmaAldrich (St. Quentin-Fallavier, France). [Glycine-2-3H]glutathione (44.8 Ci/mmol) was from New England Nuclear (Boston, MA). Ultrasphere ODS 5-Am (4.6 250 mm) column was supplied by Beckman (Margency, France). 2.2. Incubations GSH (25 mM) and ribose (25, 50, 125, 250 or 375 mM), in the presence or absence of [3H]GSH, were dissolved into 100 mM sodium phosphate buffer, pH 7.4, or 10 mM ammonium acetate, pH 6.35 for MS experiments. The mixture was sterilized by filtration (0.20 Am) and was then incubated under nitrogen in dark at 37 jC in sterile conditions. 2.3. HPLC separation The HPLC system consisted of a sample injector (100 Al, Rheodyne model 7125) and a Chromatofield pump model 501 (Chateauneuf Les Martigues, France). The resulting incubation mixture was injected into an ODS column at room temperature and isocratically eluted at 1 ml/min with 100 mM sodium phosphate buffer adjusted to pH 2.7 with phosphoric acid and filtered (0.45 Am) as a modification of the technique described by Rodriguez-Ariza et al. [16]. The products were visualized by their absorbance at 200 nm (Shimadzu SPD-6AV spectrophotometric detector) and the radioactive molecules were counted using a HPLC radioactivity monitor (Berthold LB506-C1, La Garenne-Colombes, France). Individual peaks were collected for further analyses.
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
2.4. MS The mass spectrometer used was a Finnigan TSQ 700 triple quadrupole equipped with an electrospray ionization source (ESI II). Data acquisition was performed on a DEC station 2100 running under Ultrix 4.4 (Digital Equipment) using the Finnigan software package ICIS2. The samples dissolved in a 10 mM ammonium acetate buffer (pH 6.35)
127
were infused at 5 ml/min using a Harvard 22 syringe pump. The transfer capillary was heated to 250 jC and the spray voltage was set at 4.2 kV. Mass spectra were acquired in positive ion mode by scanning from m/z 50 to m/z 800 Da within 1 s. For tandem mass spectrometry (MS/MS) experiments, daughter ion spectra were obtained at 20 eV in the laboratory frame using argon as the collision gas set at a pressure of 1 mTorr.
Fig. 2. Chromatograms of the incubation mixture of reduced glutathione with ribose. (A) A representative HPLC radioactive profile obtained after a 24-h incubation period of [3H]GSH (25 mM) and ribose (250 mM) under nitrogen in dark at 37 jC in sterile conditions in a sodium phosphate buffer (100 mM, pH 7.4). Samples were injected into 4.6 250-mm ODS column and were eluted with 100 mM sodium phosphate, pH 2.7 at a flow rate of 1 ml/min. GSH and the reaction product were eluted at 5.5 and 9 min, respectively. (B) HPLC radioactive profile obtained after a 24-h incubation period of [3H]GSH (25 mM) without ribose in the same conditions.
128
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
2.5. Determinations of thiols and GSH The concentration of thiols was determined according to a spectrophotometric assay based on 2,4-dithio-nitrobenzoic acid (Ellman’s reagent) [17]. The GSH content was measured by the HPLC technique of Neuschwander-Tetri and Roll [18] with fluorimetric detection (excitation and emission E of 340 and 420 nm, respectively). This method uses ophthalaldehyde, which specifically reacts both with the primary amine and the thiol of GSH. 2.6. Cytosolic GPx activity assay The cGPx activity was measured according to the method developed by Paglia and Valentine [19] with a few modifications. The reaction mixture (1.65 ml) consisted of 0.1-ml cGPx (10 U/ml in 50 mM Tris – HCl, 1 mM 1,4-dithioerythritol), 0.14 mM NADPH, 0.02 mM GSH and 0.6 U/ml GSSG-reductase in 50 mM Tris –HCl, 0.1 mM EDTA buffer at 37 jC. The reaction was started by addition of 50-Al tertbutyl hydroperoxide (6.6 mM) as substrate for cGPx. NADPH disappearance was monitored spectrophotometrically at 340 nm. cGPx activity was measured according to the same conditions with 0.02 mM N-ribosyl-1-glutathione instead of GSH. cGPx activity is expressed as U/Ag of enzyme where 1 U represents 1 nmol NADPH reduced/ min.
3. Results 3.1. Reaction product between GSH and ribose We conducted incubations with 250 mM ribose and 25 mM GSH in the presence of [3H]GSH under nitrogen in dark at 37 jC for 24 h. Then, aliquots of the mixture were analysed by HPLC and elution products were detected by their UV (200 nm) absorbance and their radioactivity. GSH was eluted between 5 and 6 min and another compound was eluted between 9 and 10 min (Fig. 2A), ribose being eluted in the void volume. In contrast, no peak was observed between 9 and 10 min when GSH was incubated alone under the same incubation conditions (Fig. 2B), suggesting that the compound eluted after GSH was a reaction product between GSH and ribose. 3.2. Effects of [ribose]/[GSH] ratios and incubation time on the reaction product Fig. 3A shows the extent formation of the reaction product at various [ribose]/[GSH] ratios of 1 to 15 (corresponding to ribose concentrations of 25 to 375 mM, respectively, and GSH of 25 mM) under nitrogen in dark at 37 jC for 24 h. When the [ribose]/[GSH] ratio increased, the percentage of this reaction product increased almost proportionally until the ratio 5. It is noteworthy
Fig. 3. (A) Effect of ribose and GSH concentration ratios on GSH modification. GSH and ribose were coincubated under nitrogen in dark at 37 jC in sterile conditions in 100 mM sodium phosphate buffer pH 7.4. Ribose/GSH ratios of 1, 2, 5, 10 and 15, corresponding to ribose concentrations of 25, 50, 125, 250 and 375 mM, respectively, and 25 mM GSH were used. Compounds were separated by HPLC as stated in Methods and detected by their UV (200 nm) absorbance. Values are percent of reaction product formed after 24-h incubation. Each point represents a duplicate measurement. (B) Effect of incubation time on GSH modification. GSH (25 mM) and ribose (250 mM) were incubated and analyzed in conditions described above. At the times indicated, compounds were separated by HPLC and detected by their UV (200 nm) absorbance. Values are percent of reaction product formed. This single trace is representative of separate experiments.
that it could be already detected at a ratio as lower as 1. The [ribose]/[GSH] ratio of 10 which allowed to obtain about 20% of reaction product was used for further experiments. To determine the time course of the adduct formation, incubations were conducted with the [ribose]/[GSH] ratio of 10 under nitrogen in dark from 2 to 30 h, at 37 jC. Aliquots were removed at various time intervals and the different products were separated by HPLC. The adduct formation increased almost linearly until 8 h and then increased more slowly later on, roughly attaining a maximum at 24 h (Fig. 3B). Within the first 8 h, a rate constant of 0.14 10 3 mM 1 per hour could be calculated
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
129
Fig. 4. Electrospray ionization mass spectrum of the adduct due to GSH and ribose incubation. Details on mass spectrometry are given in Methods. GSH appeared as the ion [M + H]+ (m/z 308 Th). Ribose appeared as the ion [M + NH4]+ (m/z 168 Th) and the presumed condensation product at m/z 440 Th as the [M + H]+ ion.
according to Bunn and Higgins [20]. After 24-h incubation, in the presence of [3H]GSH, the percentage of the radioactive product reached about 20% of total radioactivity (21.2 F 3.7%, mean F S.D., n = 5). Fig. 2A shows a representative HPLC radioactive profile obtained in these conditions.
3.3. Structural analysis of the reaction product: N-ribosyl1-glutathione To identify the structure of the reaction product between GSH and ribose, the incubate was analysed by electrospray MS. Fig. 4 presents the mass spectrum obtained after
Fig. 5. Daughter mass spectrum of the Amadori compound N-ribosyl-1-glutathione obtained after collision-induced dissociation of the parent ion at m/z 440 Th. The [M + H]+ ion fragment 244 indicates the cleavage of the parent ion m/z 440 Th as represented above.
130
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
electrospray ionization. The unknown compound formed was clearly detected at m/z 440 Th whereas ribose provided an ion at m/z 168 Th ([M + NH4]+) and GSH at m/z 308 Th. The ion at m/z 440 Th suggests that the unknown compound has a molecular weight of 439 Da corresponding to the condensation product between GSH and ribose. One way to make an adduct can be the reaction between the aldehyde of ribose and the N-terminus amine of GSH to form an imine product. The structure of the compound was then investigated by MS/MS. The daughter mass spectrum obtained after collision-induced dissociation of the parent ion at m/z 440 Th is presented in Fig. 5. The most intense daughter ion at m/z 244 Th suggests that this product is likely the Amadori compound, namely N-ribosyl-1-glutathione formed after condensation between ribose and GSH. The loss of successive water molecules (m/z at 226 and 208 Th) is in agreement with this hypothesis. Based upon its mass spectrometric analysis, the structure of N-ribosyl-1-glutathione shown in Figs. 4 and 5 has been proposed. In another approach to characterize the condensation product, a specific reagent of primary amine and thiol groups was used. Less than 5% of a compound reacted with o-phthalaldehyde to produce a highly fluorescent 1alkylthio, 2-alkylisoindole (results not shown), the formation of which requiring both the amine and thiol groups [18]. This result means that more than 95% of the condensation product lacked at least one of the two reactive groups required, namely amine or thiol. Moreover, the quantity of thiols in the condensation product determined by the Ellman’s method [17] (189 F 12.7 nmol, mean F S.D.; n = 3) was almost equal to the amount of the product calculated from the specific radioactivity of [3H]GSH (199 F 9.0 nmol, mean F S.D.; n = 3). Altogether these results indicated that the condensation between ribose and GSH likely involved the primary amine but not the thiol group of the tripeptide. 3.4. cGPX activity using N-ribosyl-1-glutathione as cosubstrate We used N-ribosyl-1-glutathione purified by HPLC as a cosubstrate of cGPx to determine whether glycation of GSH altered its cosubstrate property. Erythrocyte cGPx activity
Table 1 cGPx activity Cosubstrate
cGPx specific activity (U/Ag)
GSH N-ribosyl-1-GSH
20.38 F 0.15 1.16 F 0.07
cGPx activity was determined spectrophotometrically as detailed in Methods. GSH or N-ribosyl-1-glutathione was used as cosubstrate (0.02 mM) in the presence of tert-butyl hydroperoxide (80 AM), NADPH (0.14 mM), GSSG reductase (0.6 U/ml) and cGPx (0.6 U/ml). Data represent means F range of two independent experiments.
was measured by an adaptation of the Paglia and Valentine method [19]. As a matter of fact, glycated GSH appeared to be 18-fold less utilized than GSH (Table 1) and it could hardly be a cosubstrate of cGPx although it keeps its thiol group.
4. Discussion We studied the reaction of GSH with ribose in order to identify possible condensation products. The direct role of ribose in the in vivo glycation has not been established, in part because of difficulties to identify the structures involved either in vivo or in vitro. Two main reasons led us to choose ribose as a reducing sugar. First, this carbohydrate is known as a potent reducing pentose (much more reactive than glucose) due to its higher concentration in the linear openchain aldehyde form [20 – 22]. Consequently, the maximum rate of glycation can be obtained with reduced incubation time in vitro [23]. Second, ribose is a possible precursor of a fluorescent cross-linked molecule named pentosidine found in vivo [24,25], which corroborates its high reactivity as a glycation agent. In diabetes, this product was increased indeed by 2.5-fold compared to healthy control [26]. In addition, the enhancement of the pentose phosphate pathway and accelerated breakdown of ribonucleotides have been reported in diabetes, which makes likely an increased ribose mass in cells from these patients [27,28], and makes relevant its possible conjugation with GSH that is a major peptide in cell cytosol [1]. Previous works have suggested that spontaneous formation of Amadori products depends on time [3] as well as on reagent concentrations (carbohydrate and protein) [29]. The rate of reaction initially increased with ribose concentration but it was known that at high concentration (500 mM) the rate of reaction levelled off and then significantly decreased [22]. Thus the highest concentration of ribose used in our investigations was 375 mM, whereas the maximum of the reaction product was obtained with 250 mM corresponding to [ribose]/[GSH] ratio of 10. In the latter conditions, a plateau could be obtained from 24 h. In addition, the rate constant value was classically observed for the rate of condensation of monosaccharides with proteins [20]. Then 24-h incubation periods and a [ribose]/[GSH] ratio of 10 were used in further studies. Mass spectrometric analyses as well as biochemical analyses allow to determine the structure of the reaction product as an Amadori compound. Nonenzymatic binding of carbohydrates to proteins is a common biological phenomenon that is increased in diabetes in which sugar levels are elevated. The resultant product is likely to be an imine structure resulting from the condensation of the amino group with ribose which is rearranged into Amadori product as classically observed with proteins. The presence of carbohydrate on GSH may have profound effects upon its properties, e.g. alteration in its
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
interaction with enzymes. cGPx (EC 1.11.1.9) is an enzyme of intracellular defence system against lipid peroxidation. This selenoenzyme catalyses the reduction of hydroperoxides by using GSH as specific cosubstrate. GSH reacts with the active site selenium by its thiol moiety. Other groups of GSH bind electrostatically to amino acids of cGPx. In particular, hydrogen bridge is built between the amino group of the g-glutamyl residue of GSH and the carbonyl of a glycine residue [30]. Although it keeps its thiol group, the adduct cannot substitute GSH. Other interactions seem to be determinant factors in the cosubstrate reactivity. This is consistent with the data of Flohe´ [31] showing that even minor alterations of the GSH molecule result in a considerable decrease in reaction rates. cGPx is a true glutathione peroxidase, strictly speaking [31], whereas others enzymes of the GPx family such as phospholipid hydroperoxide glutathione peroxidase show a lower specificity to GSH [30]. It then appears that N-ribosyl-1-glutathione fails to be a cosubstrate for cGPx. Considering the importance of GSH in several enzyme activities and for the cell maintenance of redox status, the condensation of GSH with carbohydrates could contribute to the loss of GSH in diabetic subjects, and subsequently favour the oxidative stress associated with the disease. In vivo, the impairment on the lipoxygenase-glutathione peroxidase cascade due to carbohydrate-GSH adducts could increase the level of hydroperoxides and subsequently accelerate the whole glycation process. This would favour the development of diabetic complications.
Acknowledgements The authors thank INSERM and the Re´gion Rhoˆne-Alpes for financial support. We also thank Dr. N. Wiernsperger and Prof. C. Vial for helpful discussions.
[8] [9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17] [18]
[19]
[20] [21] [22]
References [1] D.J. Reed, Glutathione: toxicological implications, Annu. Rev. Pharmacol. Toxicol. 30 (1990) 603 – 631. [2] A. Meister, Glutathione metabolism and its selective modification, J. Biol. Chem. 263 (1988) 17205 – 17208. [3] M.P. Cohen (Ed.), Diabetes and Protein Glycosylation, Springer-Verlag, New York, 1986. [4] R.M. Bookchin, P.M. Gallop, Structure of hemoglobin A1c: nature of the N-terminal beta chain blocking group, Biochem. Biophys. Res. Commun. 32 (1968) 85 – 93. [5] R.J. Koenig, C.M. Peterson, R.L. Jones, C. Saudek, M. Lehrman, A. Cerami, Correlation of glucose regulation and hemoglobin A1c in diabetes mellitus, N. Engl. J. Med. 295 (1976) 417 – 420. [6] S. Lertsiri, M. Shiraishi, T. Miyazawa, Identification of deoxy-Dfructosyl phosphatidylethanolamine as a non-enzymic glycation product of phosphatidylethanolamine and its occurrence in human blood plasma and red blood cells, Biosci. Biotechnol. Biochem. 62 (1998) 893 – 901. [7] T. Adachi, H. Ohta, K. Hirano, K. Hayashi, S.L. Marklund, Non-
[23]
[24]
[25]
[26]
[27]
131
enzymic glycation of human extracellular superoxide dismutases, Biochem. J. 279 (1991) 263 – 267. R. Blakytny, J. Harding, Glycation (non-enzymic glycosylation) inactivates glutathione reductase, Biochem. J. 288 (1992) 303 – 307. A. Muruganandam, M. Tannous, B. Mutus, ELISA for in vivo assessment of nonenzymatically glycated platelet glutathione peroxidase, Clin. Biochem. 27 (1994) 293 – 298. J.V. Hunt, M.A. Bottoms, M.J. Mitchinson, Oxidative alterations in the experimental glycation model of diabetes mellitus are due to protein – glucose adduct oxidation, Biochem. J. 291 (1993) 529 – 535. C.G. Schalkwijk, N. Ligtvoet, H. Twaalfhoven, A. Jager, H.G. Blauawgeers, R.O. Schlingemann, L. Tarnow, H.H. Parving, C.D. Stehouwer, V.W. Van Hinsbergh, Amadori albumin in type 1 diabetic patients: correlation with markers of endothelial function, association with diabetic nephropathy, and localization in retinal capillaries, Diabetes 48 (1999) 2446 – 2453. E.A. Friedman, Advanced glycosylated end products and hyperglycemia in the pathogenesis of diabetic complications, Diabetes Care 22 (1999) B65 – B71. D. Guigliano, A. Ceriello, G. Paolisso, Diabetes mellitus, hypertension and cardiovascular disease: which role for oxidative stress? Metabolism 44 (1995) 363 – 368. G. Thomas, V. Skrinska, F.V. Lucas, O.P. Schumacher, Platelet glutathione and thromboxane synthesis in diabetes, Diabetes 34 (1985) 951 – 954. M.D. Coleman, C. Rustioni, Resistance to glutathione depletion in diabetic and non-diabetic human erythrocytes in vitro, J. Pharm. Pharmacol. 51 (1999) 21 – 25. A. Rodriguez-Ariza, F. Toribio, J. Lopez-Barea, Rapid determination of glutathione status in fish liver using high-performance liquid chromatography and electrochemical detection, J. Chromatogr., B Biomed. Sci. Appl. 656 (1994) 311 – 318. G.L. Ellman, Tissue sulfhydryl groups, Arch. Biochem. Biophys. 82 (1959) 70 – 77. B.A. Neuschwander-Tetri, F.J. Roll, Glutathione measurement by high-performance liquid chromatography separation and fluorimetric detection of the glutathione-orthophtalaldehyde adduct, Anal. Biochem. 179 (1989) 236 – 241. D.E. Paglia, W.N. Valentine, Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase, J. Lab. Clin. Med. 70 (1967) 158 – 169. H.F. Bunn, P.J. Higgins, Reaction of monosaccharides with proteins: possible evolutionary significance, Science 213 (1981) 222 – 224. H.T. Beswick, J.J. Harding, Aldehydes or dicarbonyls in non-enzymic glycosylation of proteins, Biochem. J. 226 (1985) 385 – 389. R.J. Khalifah, P. Todd, A. Ashley Booth, S.X. Yang, J.D. Mott, B.G. Hudson, Kinetics of nonenzymatic glycation of ribonuclease A leading to advanced glycation end products. Paradoxical inhibition by ribose leads to facile isolation of protein intermediate for rapid post-amadori studies, Biochemistry 35 (1996) 4645 – 4654. A.J. Bailey, T.J. Sims, N.C. Avery, E.P. Halligan, Non-enzymatic glycation of fibrous collagen: reaction products of glucose and ribose, Biochem. J. 305 (1995) 385 – 390. D.R. Sell, V.M. Monnier, Structure elucidation of a senescence crosslink from human extracellular matrix, J. Biol. Chem. 264 (1989) 21597 – 21602. S.K. Grandhee, V.M. Monnier, Mechanism of formation of the maillard protein cross-link pentosidine, J. Biol. Chem. 266 (1991) 11649 – 11653. P. Odetti, J. Fogarty, D.R. Sell, V.M. Monnier, Chromatographic quantitation of plasma and erythrocyte pentosidine in diabetic and uremic subjects, Diabetes 41 (1992) 153 – 159. A. Kashiwagi, Y. Nishio, T. Asahina, M. Ikebuchi, N. Harada, Y. Tanaka, N. Takahara, H. Taki, T. Obata, H. Hidaka, Y. Saeki, R. Kikkawa, Pyruvate improves deleterious effects of high glucose on activation of pentose phosphate pathway and glutathione redox cycle in endothelial cells, Diabetes 46 (1997) 2088 – 2095.
132
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
[28] D.R. Sell, V.M. Monnier, End-stage renal disease and diabetes catalyze the formation of a pentose-derived crosslink from aging human collagen, J. Clin. Invest. 85 (1990) 380 – 384. [29] E. Lamb, R. Mainwaring-Burton, A. Dawnay, Effect of protein concentration on the formation of glycated albumin and fructosamine, Clin. Chem. 37 (1991) 2138 – 2139.
[30] F. Ursini, M. Maiorino, R. Brigelius-Flohe´, K.D. Aumann, A. Roveri, D. Schomburg, L. Flohe´, Diversity of glutathione peroxidases, Methods Enzymol. 252 (1995) 38 – 53. [31] L. Flohe´, The selenoprotein glutathione peroxidase, in: D. Dolphin, R. Poulson, O. Avramovic (Eds.), Glutathione: Chemical, Biochemical and Medical Aspects, Wiley, New-York, 1989, pp. 643 – 731.
Covalent coupling of reduced glutathione with ribose: loss of cosubstrate ability to glutathione peroxidase Caroline Januel a, Laurent B. Fay b, Daniel Ruggiero c, Michel Lagarde a, Evelyne Ve´ricel a,* a
INSERM U 352, Biochimie et Pharmacologie, INSA-Lyon, Baˆt. Louis Pasteur, 69621 Villeurbanne Cedex, France b Nestec Ltd, Nestle´ Research Centre, Vers chez les Blancs, P.O. Box 44 1000 Lausanne, 26, Switzerland c MERCK-INSERM U 352, INSA-Lyon, Baˆt. Louis Pasteur, 69621 Villeurbanne, France Received 1 October 2002; received in revised form 2 December 2002; accepted 4 December 2002
Abstract Glycation (nonenzymatic glycosylation of proteins) is known to be increased as a result of hyperglycaemia in diabetes. Moreover, cell glutathione concentration has been found to be lower in diabetics and such depletion may impair the cell defence against toxic radical species. Ribose being a potent reducing sugar expected to be increased in cells of diabetics where the pentose phosphate pathway is enhanced, its putative condensation with glutathione was investigated. Reduced glutathione (GSH) was incubated with ribose and the structure of the resultant product was assessed by mass spectrometry, as well as the measurement of its remaining thiol group. A covalent reaction clearly occurred between the reducing sugar and GSH, to give an adduct named N-ribosyl-1-glutathione. This adduct appears to be the Amadori product resulting from the condensation of the primary amine group of GSH with the aldehyde group of ribose. Interestingly, the adduct could not be used as a proper substrate by glutathione peroxidase although it keeps its thiol group. We conclude that the coupling of GSH with a monosaccharide such as ribose might contribute to the decreased cell GSH and glutathione peroxidase activity observed in diabetics. D 2002 Elsevier Science B.V. All rights reserved. Keywords: Amadori; Diabetes; Glutathione; Glycation; Mass spectrometry; Ribose
1. Introduction Mammalian cells contain high concentrations of glutathione (g-glutamyl-cysteinyl-glycine). It is the prevalent nonprotein thiol and the most abundant low molecular weight peptide (usually in the millimolar range) [1]. Glutathione is mainly found in its reduced form (GSH) and, in much smaller amounts, in its oxidized one. GSH, a major endogenous protective system, is involved in the conjugation of several types of compounds, protects cells from the toxic effects of reactive oxygen species and maintains cellular redox status
Abbreviations: cGPx, cytosolic glutathione peroxidase; EDTA, ethylenediaminetetraacetic acid; GSH, reduced glutathione; GSSG, oxidized glutathione; HPLC, high-performance liquid chromatography; MS, mass spectrometry; MS/MS, tandem mass spectrometry; NADPH, nicotinamide adenine dinucleotide phosphate (reduced form) * Corresponding author. Tel.: +33-4-72-43-84-79; fax: +33-4-72-4385-24. E-mail address: [email protected]. (E. Ve´ricel).
[2]. Finally, GSH is a cosubstrate of several enzymes including GSH peroxidase (GPx). Nonenzymatic glycosylation (glycation) results from the chemical reaction between carbohydrates and proteins. The reaction is accelerated in response to hyperglycaemia in diabetes and is involved in the pathogenesis of its chronic complications [3]. The very early products of the glycation process such as the Schiff base and Amadori products result from the covalent addition of glucose (or other reducing carbohydrates) to protein amino groups (Fig. 1). The first protein demonstrated to be glycated in vivo was haemoglobin [4], and the fraction of glycated haemoglobin increased from 5% in normal people to 10% in diabetic patients. Indeed, it is considered that the level of glycated haemoglobin also reflects the blood glucose concentration of such patients [5]. Since this discovery, glycation of a number of proteins (collagen, albumin, crystalline, etc.) [6] and several enzymes (erythrocyte superoxide dismutase, erythrocyte purine nucleoside phosphorylase, platelet GPx, etc.) [7– 9] has been reported. Such a glycation often impairs the function of proteins and contributes to inhibition of enzyme
0304-4165/02/$ - see front matter D 2002 Elsevier Science B.V. All rights reserved. doi:10.1016/S0304-4165(02)00525-1
126
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
Fig. 1. Early products in protein condensation with glucose.
activity [7]. Amadori products may be transformed in a variety of compounds until advanced glycosylation endproducts (AGE). The AGE formation is known to be affected by different factors, including metal ions which catalyze both the glycoxidation and the conversion of Amadori products [10]. Both early glycation Amadorimodified protein [11] and AGE [12] formation were found increased in diabetes. Moreover, in diabetes, high glucose concentrations lead to both formation of toxic reactive oxygen species [13] and alteration in cell defence systems such as GSSG-reductase [8], GPx [9] and superoxide dismutase [6]. Hence, the protection against toxic reactive species is lower in diabetics and GSH maintenance is impaired [14,15]. One working hypothesis to explain the decrease of GSH in diabetes could be the glycation of GSH. To determine whether GSH could be first condensed with a carbohydrate, it was incubated with ribose, a potent glycating agent assumed to be increased in diabetes. The condensation product was isolated by HPLC and its structure determined by mass spectrometry (MS) and named N-ribosyl-1-glutathione. Finally, N-ribosyl-1-glutathione was tested as a cosubstrate for cytosolic GPx (cGPx) instead of GSH.
2. Methods 2.1. Materials Glutathione (reduced form), D( )-ribose, NADPH, bovine erythrocyte cGPx, GSSG-reductase, tert-butyl hydro-
peroxide and other reagents were obtained from SigmaAldrich (St. Quentin-Fallavier, France). [Glycine-2-3H]glutathione (44.8 Ci/mmol) was from New England Nuclear (Boston, MA). Ultrasphere ODS 5-Am (4.6 250 mm) column was supplied by Beckman (Margency, France). 2.2. Incubations GSH (25 mM) and ribose (25, 50, 125, 250 or 375 mM), in the presence or absence of [3H]GSH, were dissolved into 100 mM sodium phosphate buffer, pH 7.4, or 10 mM ammonium acetate, pH 6.35 for MS experiments. The mixture was sterilized by filtration (0.20 Am) and was then incubated under nitrogen in dark at 37 jC in sterile conditions. 2.3. HPLC separation The HPLC system consisted of a sample injector (100 Al, Rheodyne model 7125) and a Chromatofield pump model 501 (Chateauneuf Les Martigues, France). The resulting incubation mixture was injected into an ODS column at room temperature and isocratically eluted at 1 ml/min with 100 mM sodium phosphate buffer adjusted to pH 2.7 with phosphoric acid and filtered (0.45 Am) as a modification of the technique described by Rodriguez-Ariza et al. [16]. The products were visualized by their absorbance at 200 nm (Shimadzu SPD-6AV spectrophotometric detector) and the radioactive molecules were counted using a HPLC radioactivity monitor (Berthold LB506-C1, La Garenne-Colombes, France). Individual peaks were collected for further analyses.
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
2.4. MS The mass spectrometer used was a Finnigan TSQ 700 triple quadrupole equipped with an electrospray ionization source (ESI II). Data acquisition was performed on a DEC station 2100 running under Ultrix 4.4 (Digital Equipment) using the Finnigan software package ICIS2. The samples dissolved in a 10 mM ammonium acetate buffer (pH 6.35)
127
were infused at 5 ml/min using a Harvard 22 syringe pump. The transfer capillary was heated to 250 jC and the spray voltage was set at 4.2 kV. Mass spectra were acquired in positive ion mode by scanning from m/z 50 to m/z 800 Da within 1 s. For tandem mass spectrometry (MS/MS) experiments, daughter ion spectra were obtained at 20 eV in the laboratory frame using argon as the collision gas set at a pressure of 1 mTorr.
Fig. 2. Chromatograms of the incubation mixture of reduced glutathione with ribose. (A) A representative HPLC radioactive profile obtained after a 24-h incubation period of [3H]GSH (25 mM) and ribose (250 mM) under nitrogen in dark at 37 jC in sterile conditions in a sodium phosphate buffer (100 mM, pH 7.4). Samples were injected into 4.6 250-mm ODS column and were eluted with 100 mM sodium phosphate, pH 2.7 at a flow rate of 1 ml/min. GSH and the reaction product were eluted at 5.5 and 9 min, respectively. (B) HPLC radioactive profile obtained after a 24-h incubation period of [3H]GSH (25 mM) without ribose in the same conditions.
128
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
2.5. Determinations of thiols and GSH The concentration of thiols was determined according to a spectrophotometric assay based on 2,4-dithio-nitrobenzoic acid (Ellman’s reagent) [17]. The GSH content was measured by the HPLC technique of Neuschwander-Tetri and Roll [18] with fluorimetric detection (excitation and emission E of 340 and 420 nm, respectively). This method uses ophthalaldehyde, which specifically reacts both with the primary amine and the thiol of GSH. 2.6. Cytosolic GPx activity assay The cGPx activity was measured according to the method developed by Paglia and Valentine [19] with a few modifications. The reaction mixture (1.65 ml) consisted of 0.1-ml cGPx (10 U/ml in 50 mM Tris – HCl, 1 mM 1,4-dithioerythritol), 0.14 mM NADPH, 0.02 mM GSH and 0.6 U/ml GSSG-reductase in 50 mM Tris –HCl, 0.1 mM EDTA buffer at 37 jC. The reaction was started by addition of 50-Al tertbutyl hydroperoxide (6.6 mM) as substrate for cGPx. NADPH disappearance was monitored spectrophotometrically at 340 nm. cGPx activity was measured according to the same conditions with 0.02 mM N-ribosyl-1-glutathione instead of GSH. cGPx activity is expressed as U/Ag of enzyme where 1 U represents 1 nmol NADPH reduced/ min.
3. Results 3.1. Reaction product between GSH and ribose We conducted incubations with 250 mM ribose and 25 mM GSH in the presence of [3H]GSH under nitrogen in dark at 37 jC for 24 h. Then, aliquots of the mixture were analysed by HPLC and elution products were detected by their UV (200 nm) absorbance and their radioactivity. GSH was eluted between 5 and 6 min and another compound was eluted between 9 and 10 min (Fig. 2A), ribose being eluted in the void volume. In contrast, no peak was observed between 9 and 10 min when GSH was incubated alone under the same incubation conditions (Fig. 2B), suggesting that the compound eluted after GSH was a reaction product between GSH and ribose. 3.2. Effects of [ribose]/[GSH] ratios and incubation time on the reaction product Fig. 3A shows the extent formation of the reaction product at various [ribose]/[GSH] ratios of 1 to 15 (corresponding to ribose concentrations of 25 to 375 mM, respectively, and GSH of 25 mM) under nitrogen in dark at 37 jC for 24 h. When the [ribose]/[GSH] ratio increased, the percentage of this reaction product increased almost proportionally until the ratio 5. It is noteworthy
Fig. 3. (A) Effect of ribose and GSH concentration ratios on GSH modification. GSH and ribose were coincubated under nitrogen in dark at 37 jC in sterile conditions in 100 mM sodium phosphate buffer pH 7.4. Ribose/GSH ratios of 1, 2, 5, 10 and 15, corresponding to ribose concentrations of 25, 50, 125, 250 and 375 mM, respectively, and 25 mM GSH were used. Compounds were separated by HPLC as stated in Methods and detected by their UV (200 nm) absorbance. Values are percent of reaction product formed after 24-h incubation. Each point represents a duplicate measurement. (B) Effect of incubation time on GSH modification. GSH (25 mM) and ribose (250 mM) were incubated and analyzed in conditions described above. At the times indicated, compounds were separated by HPLC and detected by their UV (200 nm) absorbance. Values are percent of reaction product formed. This single trace is representative of separate experiments.
that it could be already detected at a ratio as lower as 1. The [ribose]/[GSH] ratio of 10 which allowed to obtain about 20% of reaction product was used for further experiments. To determine the time course of the adduct formation, incubations were conducted with the [ribose]/[GSH] ratio of 10 under nitrogen in dark from 2 to 30 h, at 37 jC. Aliquots were removed at various time intervals and the different products were separated by HPLC. The adduct formation increased almost linearly until 8 h and then increased more slowly later on, roughly attaining a maximum at 24 h (Fig. 3B). Within the first 8 h, a rate constant of 0.14 10 3 mM 1 per hour could be calculated
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
129
Fig. 4. Electrospray ionization mass spectrum of the adduct due to GSH and ribose incubation. Details on mass spectrometry are given in Methods. GSH appeared as the ion [M + H]+ (m/z 308 Th). Ribose appeared as the ion [M + NH4]+ (m/z 168 Th) and the presumed condensation product at m/z 440 Th as the [M + H]+ ion.
according to Bunn and Higgins [20]. After 24-h incubation, in the presence of [3H]GSH, the percentage of the radioactive product reached about 20% of total radioactivity (21.2 F 3.7%, mean F S.D., n = 5). Fig. 2A shows a representative HPLC radioactive profile obtained in these conditions.
3.3. Structural analysis of the reaction product: N-ribosyl1-glutathione To identify the structure of the reaction product between GSH and ribose, the incubate was analysed by electrospray MS. Fig. 4 presents the mass spectrum obtained after
Fig. 5. Daughter mass spectrum of the Amadori compound N-ribosyl-1-glutathione obtained after collision-induced dissociation of the parent ion at m/z 440 Th. The [M + H]+ ion fragment 244 indicates the cleavage of the parent ion m/z 440 Th as represented above.
130
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
electrospray ionization. The unknown compound formed was clearly detected at m/z 440 Th whereas ribose provided an ion at m/z 168 Th ([M + NH4]+) and GSH at m/z 308 Th. The ion at m/z 440 Th suggests that the unknown compound has a molecular weight of 439 Da corresponding to the condensation product between GSH and ribose. One way to make an adduct can be the reaction between the aldehyde of ribose and the N-terminus amine of GSH to form an imine product. The structure of the compound was then investigated by MS/MS. The daughter mass spectrum obtained after collision-induced dissociation of the parent ion at m/z 440 Th is presented in Fig. 5. The most intense daughter ion at m/z 244 Th suggests that this product is likely the Amadori compound, namely N-ribosyl-1-glutathione formed after condensation between ribose and GSH. The loss of successive water molecules (m/z at 226 and 208 Th) is in agreement with this hypothesis. Based upon its mass spectrometric analysis, the structure of N-ribosyl-1-glutathione shown in Figs. 4 and 5 has been proposed. In another approach to characterize the condensation product, a specific reagent of primary amine and thiol groups was used. Less than 5% of a compound reacted with o-phthalaldehyde to produce a highly fluorescent 1alkylthio, 2-alkylisoindole (results not shown), the formation of which requiring both the amine and thiol groups [18]. This result means that more than 95% of the condensation product lacked at least one of the two reactive groups required, namely amine or thiol. Moreover, the quantity of thiols in the condensation product determined by the Ellman’s method [17] (189 F 12.7 nmol, mean F S.D.; n = 3) was almost equal to the amount of the product calculated from the specific radioactivity of [3H]GSH (199 F 9.0 nmol, mean F S.D.; n = 3). Altogether these results indicated that the condensation between ribose and GSH likely involved the primary amine but not the thiol group of the tripeptide. 3.4. cGPX activity using N-ribosyl-1-glutathione as cosubstrate We used N-ribosyl-1-glutathione purified by HPLC as a cosubstrate of cGPx to determine whether glycation of GSH altered its cosubstrate property. Erythrocyte cGPx activity
Table 1 cGPx activity Cosubstrate
cGPx specific activity (U/Ag)
GSH N-ribosyl-1-GSH
20.38 F 0.15 1.16 F 0.07
cGPx activity was determined spectrophotometrically as detailed in Methods. GSH or N-ribosyl-1-glutathione was used as cosubstrate (0.02 mM) in the presence of tert-butyl hydroperoxide (80 AM), NADPH (0.14 mM), GSSG reductase (0.6 U/ml) and cGPx (0.6 U/ml). Data represent means F range of two independent experiments.
was measured by an adaptation of the Paglia and Valentine method [19]. As a matter of fact, glycated GSH appeared to be 18-fold less utilized than GSH (Table 1) and it could hardly be a cosubstrate of cGPx although it keeps its thiol group.
4. Discussion We studied the reaction of GSH with ribose in order to identify possible condensation products. The direct role of ribose in the in vivo glycation has not been established, in part because of difficulties to identify the structures involved either in vivo or in vitro. Two main reasons led us to choose ribose as a reducing sugar. First, this carbohydrate is known as a potent reducing pentose (much more reactive than glucose) due to its higher concentration in the linear openchain aldehyde form [20 – 22]. Consequently, the maximum rate of glycation can be obtained with reduced incubation time in vitro [23]. Second, ribose is a possible precursor of a fluorescent cross-linked molecule named pentosidine found in vivo [24,25], which corroborates its high reactivity as a glycation agent. In diabetes, this product was increased indeed by 2.5-fold compared to healthy control [26]. In addition, the enhancement of the pentose phosphate pathway and accelerated breakdown of ribonucleotides have been reported in diabetes, which makes likely an increased ribose mass in cells from these patients [27,28], and makes relevant its possible conjugation with GSH that is a major peptide in cell cytosol [1]. Previous works have suggested that spontaneous formation of Amadori products depends on time [3] as well as on reagent concentrations (carbohydrate and protein) [29]. The rate of reaction initially increased with ribose concentration but it was known that at high concentration (500 mM) the rate of reaction levelled off and then significantly decreased [22]. Thus the highest concentration of ribose used in our investigations was 375 mM, whereas the maximum of the reaction product was obtained with 250 mM corresponding to [ribose]/[GSH] ratio of 10. In the latter conditions, a plateau could be obtained from 24 h. In addition, the rate constant value was classically observed for the rate of condensation of monosaccharides with proteins [20]. Then 24-h incubation periods and a [ribose]/[GSH] ratio of 10 were used in further studies. Mass spectrometric analyses as well as biochemical analyses allow to determine the structure of the reaction product as an Amadori compound. Nonenzymatic binding of carbohydrates to proteins is a common biological phenomenon that is increased in diabetes in which sugar levels are elevated. The resultant product is likely to be an imine structure resulting from the condensation of the amino group with ribose which is rearranged into Amadori product as classically observed with proteins. The presence of carbohydrate on GSH may have profound effects upon its properties, e.g. alteration in its
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
interaction with enzymes. cGPx (EC 1.11.1.9) is an enzyme of intracellular defence system against lipid peroxidation. This selenoenzyme catalyses the reduction of hydroperoxides by using GSH as specific cosubstrate. GSH reacts with the active site selenium by its thiol moiety. Other groups of GSH bind electrostatically to amino acids of cGPx. In particular, hydrogen bridge is built between the amino group of the g-glutamyl residue of GSH and the carbonyl of a glycine residue [30]. Although it keeps its thiol group, the adduct cannot substitute GSH. Other interactions seem to be determinant factors in the cosubstrate reactivity. This is consistent with the data of Flohe´ [31] showing that even minor alterations of the GSH molecule result in a considerable decrease in reaction rates. cGPx is a true glutathione peroxidase, strictly speaking [31], whereas others enzymes of the GPx family such as phospholipid hydroperoxide glutathione peroxidase show a lower specificity to GSH [30]. It then appears that N-ribosyl-1-glutathione fails to be a cosubstrate for cGPx. Considering the importance of GSH in several enzyme activities and for the cell maintenance of redox status, the condensation of GSH with carbohydrates could contribute to the loss of GSH in diabetic subjects, and subsequently favour the oxidative stress associated with the disease. In vivo, the impairment on the lipoxygenase-glutathione peroxidase cascade due to carbohydrate-GSH adducts could increase the level of hydroperoxides and subsequently accelerate the whole glycation process. This would favour the development of diabetic complications.
Acknowledgements The authors thank INSERM and the Re´gion Rhoˆne-Alpes for financial support. We also thank Dr. N. Wiernsperger and Prof. C. Vial for helpful discussions.
[8] [9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17] [18]
[19]
[20] [21] [22]
References [1] D.J. Reed, Glutathione: toxicological implications, Annu. Rev. Pharmacol. Toxicol. 30 (1990) 603 – 631. [2] A. Meister, Glutathione metabolism and its selective modification, J. Biol. Chem. 263 (1988) 17205 – 17208. [3] M.P. Cohen (Ed.), Diabetes and Protein Glycosylation, Springer-Verlag, New York, 1986. [4] R.M. Bookchin, P.M. Gallop, Structure of hemoglobin A1c: nature of the N-terminal beta chain blocking group, Biochem. Biophys. Res. Commun. 32 (1968) 85 – 93. [5] R.J. Koenig, C.M. Peterson, R.L. Jones, C. Saudek, M. Lehrman, A. Cerami, Correlation of glucose regulation and hemoglobin A1c in diabetes mellitus, N. Engl. J. Med. 295 (1976) 417 – 420. [6] S. Lertsiri, M. Shiraishi, T. Miyazawa, Identification of deoxy-Dfructosyl phosphatidylethanolamine as a non-enzymic glycation product of phosphatidylethanolamine and its occurrence in human blood plasma and red blood cells, Biosci. Biotechnol. Biochem. 62 (1998) 893 – 901. [7] T. Adachi, H. Ohta, K. Hirano, K. Hayashi, S.L. Marklund, Non-
[23]
[24]
[25]
[26]
[27]
131
enzymic glycation of human extracellular superoxide dismutases, Biochem. J. 279 (1991) 263 – 267. R. Blakytny, J. Harding, Glycation (non-enzymic glycosylation) inactivates glutathione reductase, Biochem. J. 288 (1992) 303 – 307. A. Muruganandam, M. Tannous, B. Mutus, ELISA for in vivo assessment of nonenzymatically glycated platelet glutathione peroxidase, Clin. Biochem. 27 (1994) 293 – 298. J.V. Hunt, M.A. Bottoms, M.J. Mitchinson, Oxidative alterations in the experimental glycation model of diabetes mellitus are due to protein – glucose adduct oxidation, Biochem. J. 291 (1993) 529 – 535. C.G. Schalkwijk, N. Ligtvoet, H. Twaalfhoven, A. Jager, H.G. Blauawgeers, R.O. Schlingemann, L. Tarnow, H.H. Parving, C.D. Stehouwer, V.W. Van Hinsbergh, Amadori albumin in type 1 diabetic patients: correlation with markers of endothelial function, association with diabetic nephropathy, and localization in retinal capillaries, Diabetes 48 (1999) 2446 – 2453. E.A. Friedman, Advanced glycosylated end products and hyperglycemia in the pathogenesis of diabetic complications, Diabetes Care 22 (1999) B65 – B71. D. Guigliano, A. Ceriello, G. Paolisso, Diabetes mellitus, hypertension and cardiovascular disease: which role for oxidative stress? Metabolism 44 (1995) 363 – 368. G. Thomas, V. Skrinska, F.V. Lucas, O.P. Schumacher, Platelet glutathione and thromboxane synthesis in diabetes, Diabetes 34 (1985) 951 – 954. M.D. Coleman, C. Rustioni, Resistance to glutathione depletion in diabetic and non-diabetic human erythrocytes in vitro, J. Pharm. Pharmacol. 51 (1999) 21 – 25. A. Rodriguez-Ariza, F. Toribio, J. Lopez-Barea, Rapid determination of glutathione status in fish liver using high-performance liquid chromatography and electrochemical detection, J. Chromatogr., B Biomed. Sci. Appl. 656 (1994) 311 – 318. G.L. Ellman, Tissue sulfhydryl groups, Arch. Biochem. Biophys. 82 (1959) 70 – 77. B.A. Neuschwander-Tetri, F.J. Roll, Glutathione measurement by high-performance liquid chromatography separation and fluorimetric detection of the glutathione-orthophtalaldehyde adduct, Anal. Biochem. 179 (1989) 236 – 241. D.E. Paglia, W.N. Valentine, Studies on the quantitative and qualitative characterization of erythrocyte glutathione peroxidase, J. Lab. Clin. Med. 70 (1967) 158 – 169. H.F. Bunn, P.J. Higgins, Reaction of monosaccharides with proteins: possible evolutionary significance, Science 213 (1981) 222 – 224. H.T. Beswick, J.J. Harding, Aldehydes or dicarbonyls in non-enzymic glycosylation of proteins, Biochem. J. 226 (1985) 385 – 389. R.J. Khalifah, P. Todd, A. Ashley Booth, S.X. Yang, J.D. Mott, B.G. Hudson, Kinetics of nonenzymatic glycation of ribonuclease A leading to advanced glycation end products. Paradoxical inhibition by ribose leads to facile isolation of protein intermediate for rapid post-amadori studies, Biochemistry 35 (1996) 4645 – 4654. A.J. Bailey, T.J. Sims, N.C. Avery, E.P. Halligan, Non-enzymatic glycation of fibrous collagen: reaction products of glucose and ribose, Biochem. J. 305 (1995) 385 – 390. D.R. Sell, V.M. Monnier, Structure elucidation of a senescence crosslink from human extracellular matrix, J. Biol. Chem. 264 (1989) 21597 – 21602. S.K. Grandhee, V.M. Monnier, Mechanism of formation of the maillard protein cross-link pentosidine, J. Biol. Chem. 266 (1991) 11649 – 11653. P. Odetti, J. Fogarty, D.R. Sell, V.M. Monnier, Chromatographic quantitation of plasma and erythrocyte pentosidine in diabetic and uremic subjects, Diabetes 41 (1992) 153 – 159. A. Kashiwagi, Y. Nishio, T. Asahina, M. Ikebuchi, N. Harada, Y. Tanaka, N. Takahara, H. Taki, T. Obata, H. Hidaka, Y. Saeki, R. Kikkawa, Pyruvate improves deleterious effects of high glucose on activation of pentose phosphate pathway and glutathione redox cycle in endothelial cells, Diabetes 46 (1997) 2088 – 2095.
132
C. Januel et al. / Biochimica et Biophysica Acta 1620 (2003) 125–132
[28] D.R. Sell, V.M. Monnier, End-stage renal disease and diabetes catalyze the formation of a pentose-derived crosslink from aging human collagen, J. Clin. Invest. 85 (1990) 380 – 384. [29] E. Lamb, R. Mainwaring-Burton, A. Dawnay, Effect of protein concentration on the formation of glycated albumin and fructosamine, Clin. Chem. 37 (1991) 2138 – 2139.
[30] F. Ursini, M. Maiorino, R. Brigelius-Flohe´, K.D. Aumann, A. Roveri, D. Schomburg, L. Flohe´, Diversity of glutathione peroxidases, Methods Enzymol. 252 (1995) 38 – 53. [31] L. Flohe´, The selenoprotein glutathione peroxidase, in: D. Dolphin, R. Poulson, O. Avramovic (Eds.), Glutathione: Chemical, Biochemical and Medical Aspects, Wiley, New-York, 1989, pp. 643 – 731.