Amides Are Novel Protein Modifications Formed by Physiological Sugars
2001; Elsevier BV; Volume: 276; Issue: 45 Linguagem: Inglês
10.1074/jbc.m103557200
ISSN1083-351X
AutoresMarcus A. Glomb, Christoph Pfahler,
Tópico(s)Biochemical effects in animals
ResumoThe Maillard reaction, or nonenzymatic browning, proceeds in vivo, and the resulting protein modifications (advanced glycation end products) have been associated with various pathologies. Despite intensive research only very few structures have been established in vivo. We report here for the first time N 6-{2-[(5-amino-5-carboxypentyl)amino]-2-oxoethyl}lysine (GOLA) and N 6-glycoloyllysine (GALA) as prototypes for novel amide protein modifications produced by reducing sugars. Their identity was confirmed by independent synthesis and coupled liquid chromatography/mass spectrometry. Model reactions with N α-t-butoxycarbonyl-lysine showed that glyoxal and glycolaldehyde are immediate precursors, and reaction pathways are directly linked to N ε-carboxymethyllysine via glyoxal-imine structures. GOLA, the amide cross-link, and 1,3-bis(5-amino-5-carboxypentyl)imidazolium salt (GOLD), the imidazolium cross-link, share a common intermediate. The ratio of GOLA to GOLD is greater when glyoxal levels are low at constant lysine concentrations. GOLA and GALA formation from the Amadori product of glucose and lysine depends directly upon oxidation. With the advanced glycation end product inhibitors aminoguanidine and pyridoxamine we were able to dissect oxidative fragmentation of the Amadori product as a second mechanism of GOLA formation exactly coinciding with N ε-carboxymethyllysine synthesis. In contrast, the formation of GALA appears to depend solely upon glyoxal-imines. After enzymatic hydrolysis GOLA was found at 66 pmol/mg of brunescent lens protein. This suggests amide protein modifications as important markers of pathophysiological processes. The Maillard reaction, or nonenzymatic browning, proceeds in vivo, and the resulting protein modifications (advanced glycation end products) have been associated with various pathologies. Despite intensive research only very few structures have been established in vivo. We report here for the first time N 6-{2-[(5-amino-5-carboxypentyl)amino]-2-oxoethyl}lysine (GOLA) and N 6-glycoloyllysine (GALA) as prototypes for novel amide protein modifications produced by reducing sugars. Their identity was confirmed by independent synthesis and coupled liquid chromatography/mass spectrometry. Model reactions with N α-t-butoxycarbonyl-lysine showed that glyoxal and glycolaldehyde are immediate precursors, and reaction pathways are directly linked to N ε-carboxymethyllysine via glyoxal-imine structures. GOLA, the amide cross-link, and 1,3-bis(5-amino-5-carboxypentyl)imidazolium salt (GOLD), the imidazolium cross-link, share a common intermediate. The ratio of GOLA to GOLD is greater when glyoxal levels are low at constant lysine concentrations. GOLA and GALA formation from the Amadori product of glucose and lysine depends directly upon oxidation. With the advanced glycation end product inhibitors aminoguanidine and pyridoxamine we were able to dissect oxidative fragmentation of the Amadori product as a second mechanism of GOLA formation exactly coinciding with N ε-carboxymethyllysine synthesis. In contrast, the formation of GALA appears to depend solely upon glyoxal-imines. After enzymatic hydrolysis GOLA was found at 66 pmol/mg of brunescent lens protein. This suggests amide protein modifications as important markers of pathophysiological processes. advanced glycation end product N ε-carboxymethyllysine 1,3-bis(5-amino-5-carboxypentyl)imidazolium salt pyridoxamine bovine serum albumin N 6-{2-[(5-amino-5-carboxypentyl)amino]-2-oxoethyl}lysine N 6-glycoloyllysine [(3-hydroxy-5-hydroxymethyl-2-methyl-pyridin-4-ylmethyl)amino]acetic acid aminoguanidine high performance liquid chromatography electrospray ionization mass spectrometry heptafluorobutyric acid butoxycarbonyl 2-ammonio-6-([2-[(4-ammonio-5-oxido-5- oxopentyl)amino]-4,5-dihydro-1 H-imidazol-5-ylidene]amino)hexanoate 2-ammonio-6-([2-[(4-ammonio-5-oxido-5-oxopentyl)amino]-4-methyl-4,5-dihydro-1 H-imidazol-5-ylidene]amino)hexanoate N,N(-di(N ε-lysino))-4-methyl-imidazolium (methylglyoxal-lysine dimer) thin-layer chromatography room temperature The Maillard reaction, or nonenzymatic browning, occurs in vivo and contributes to the late complications of diabetes, atherosclerosis, and aging. Model reaction systems show that reducing sugars, such as glucose, can react with proteins or other amine structures via highly reactive intermediates in complex reaction cascades. A variety of stable end products, known as advanced glycation end products, or AGEs,1 form as a result (1Ledl F. Schleicher E. Angew. Chem. Int. Ed. Engl. 1990; 29: 565-594Crossref Scopus (684) Google Scholar). However, the few AGEs identified in vivocannot account for the extensive protein modifications found in target tissues (2Monnier V.M. Nagaraj R.H. Potero-Otin M. Glomb M.A. Elgawish A.H. Sell D.R. Friedlander M.A. Nephrol. Dial. Transplant. 1996; 11: 20-26Crossref PubMed Scopus (69) Google Scholar, 3Monnier V.M. Glomb M.A. Elgawish A. Sell D.R. Diabetes. 1996; 45 Suppl. 3: 67-72Crossref Google Scholar). Since AGEs are recognized by cellular receptors, they can also damage tissues by triggering the production of reactive oxygen species (4Schmidt A.M. Hori O. Brett J. Yan S.D. Wautier J.-L. Stern S. Aterioscler. Thromb. 1994; 14: 1521-1528Crossref PubMed Google Scholar). Elucidation of AGE structures and mechanisms of their formation will aid in the development of pharmaceutical strategies to prevent AGE-associated complications. Both α-dicarbonyl structures and oxidative radical mechanisms are thought to play an important role within the Maillard reaction (5Wolff S.P. Jiang Z.Y. Hunt J.V. Free Radic. Biol. Med. 1991; 10: 339-352Crossref PubMed Scopus (802) Google Scholar, 6Fu M.-X. Knecht K.J. Thorpe S.R. Baynes J.W. Diabetes. 1992; 41: 42-48Crossref PubMed Google Scholar). Oxygen radical reactions are directly coupled to the cross-linking of proteins and to the formation of AGEs such as N ε-carboxymethyllysine (CML). They are catalyzed by transition metal ions and can be inhibited by chelating agents. CML, the most prevalent AGE identified in vivo, is the major antigen recognized by polyclonal antibodies raised against glycated proteins (7Reddy S. Bichler J. Wells-Knecht K.J. Thorpe S.R. Baynes J.W. Biochemistry. 1995; 34: 10872-10878Crossref PubMed Scopus (469) Google Scholar). CML concentrations in collagen increase with age along with an overall increase in fluorescence, but CML concentrations rise markedly in diabetes. However, discrepancies between the extent of protein modification and fluorescence suggest that not all cross-linking modifications are due to colored, ultraviolet- or fluorescence-active structures. Thus, pentosidine, crosslines, and vesperlysine A, all of which are modifiers of proteins in vivo (8Sell D.R. Lapolla A. Odetti P. Fogarty J. Monnier V.M. Diabetes. 1992; 41: 1286-1292Crossref PubMed Google Scholar, 9Nakamura Y. Horii Y. Nishino T. Shiiki H. Sakaguchi Y. Kagoshima T. Dohi K. Makita Z. Vlassara H. Bucala R. Am. J. Pathol. 1993; 143: 1649-1656PubMed Google Scholar, 10Tessier F. Obrenovich M. Monnier V.M. J. Biol. Chem. 1999; 274: 20796-20804Abstract Full Text Full Text PDF PubMed Scopus (165) Google Scholar), might not be as important as originally supposed. Several laboratories have attempted to define the structures of potential cross-link moieties. Eble et al. (11Eble A.S. Thorpe S.R. Baynes J.W. J. Biol. Chem. 1983; 285: 9406-9412Abstract Full Text PDF Google Scholar) found that radioactive labeled N α-formyllysine was bound covalently to glycated proteins only under oxidative conditions. Major quantities of lysine were released by acid and alkaline hydrolysis; this reaction could not be inhibited by prior reduction with sodium borohydride. Thus, the acid-stable imidazolium cross-links derived from the reaction of lysine with glyoxal and methylglyoxal (1,3-bis(5-amino-5-carboxypentyl)imidazolium salt (GOLD), MOLD/imidazolysine; Refs. 12Odani H. Shinzato T. Usami J. Matsumoto Y. Brinkmann-Frye E. Baynes J.W. Maeda K. FEBS Lett. 1998; 427: 381-385Crossref PubMed Scopus (138) Google Scholar and 13Nagaraj R.H. Shipanova I.N. Faust F.M. J. Biol. Chem. 1996; 271: 19338-19345Abstract Full Text Full Text PDF PubMed Scopus (284) Google Scholar) might not be the most important physiologically, although they are the most prevalent cross-linking structures thus far identified in vivo. Lederer et al. (14Lederer M.O. Bühler H.P. Bioorg. Med. Chem. 1999; 7: 2499-2507Crossref PubMed Scopus (126) Google Scholar) incubated bovine serum albumin (BSA) with glucose under physiological conditions and reported the alternative structures GODIC, MODIC, and glucosepan. These imidazole ring-based structures cross-link protein-bound arginine and lysine residues, but they can be analyzed only after enzymatic hydrolysis because of their structural instability. Amides are novel AGEs that, like other known AGEs, result in protein cross-linking. Once we were able to synthesize GOLA or glyoxal-lysine-amide (N 6-{2-[(5-amino-5-carboxypentyl)amino]-2-oxoethyl}lysine) and GALA or glycolic acid-lysine-amide (N 6-glycoloyllysine) independently, we developed a means to detect them in modified proteins. We then confirmed their formation from physiological important sugars and showed that the underlying mechanisms were linked directly to CML. Our identification of GOLA in human lens proteins shows that these amides can form in human tissue proteins. We measured formation of GOLD in all incubations to determine the potential significance of GOLA and GALA for in vivo systems. Our experiments also explored the actions of aminoguanidine (AG) and pyridoxamine (PM), both of which are inhibitors of advanced glycation reactions in vivo (15Khalifah R.G. Baynes J.W. Hudson B.H. Biochem. Biophys. Res. Commun. 1999; 257: 251-258Crossref PubMed Scopus (252) Google Scholar). Chemicals of the highest quality available were obtained from Aldrich, Sigma, and Fluka (Neu-Ulm, Germany) unless otherwise indicated. Thin-layer chromatography (TLC) was performed on silica gel 60 F254 plates (Merck). Visualization of separated material was achieved with ninhydrin unless otherwise indicated. Preparative column chromatography was performed on silica gel 60 (63–200 μm, Merck; atmospheric pressure), and reversed phase material (Lobar RP-18, 25 × 310 mm, 40–63 μm, Merck; flow: 5 ml/min, Besta HD 2-200 pump, Wilhelmsfeld, Germany). Solvents were all of chromatographic grade. From the individual fractions, solvents were removed under reduced pressure. N α-t-Butoxycarbonyl-N ε-(1-deoxy-d-fructos-1-yl)lysine (Amadori product of d-glucose and N α-t-Boc-lysine) was synthesized according to Glomb and Monnier (16Glomb M.A. Monnier V.M. J. Biol. Chem. 1995; 270: 10017-10026Abstract Full Text Full Text PDF PubMed Scopus (517) Google Scholar). The synthesis of GOLA and GALA is outlined in Fig. 1.t-Butyl-N 6-[(benzyloxy)carbonyl]-N 2-(t-butoxycarbonyl)lysinate (2) and t-butyl-N 2-(t-butoxycarbonyl)lysinate (3) were synthesized according to Narayanan and Griffith (17Narayanan K. Griffith K. J. Med. Chem. 1994; 37: 885-887Crossref PubMed Scopus (111) Google Scholar) with major modifications. Briefly, 300 μl of HClO4(60%) was added to a solution of 3.8 g (10 mmol) of N α-t-Boc-N ε-benzyloxycarbonyl-lysine in 150 ml of t-butylacetate and stirred for 48 h at RT. Solvents were evaporated, and the resulting residue was subjected to column chromatography (silica gel; EtOAc/n-hexane, 3:7, v/v). Fractions with material having a Rf value of 0.36 were combined (TLC, same eluent), and solvents were evaporated to yield 2 as a colorless oil (0.34 g, 7.8%). 59 mg of palladium/charcoal (10%) was added to a solution of 618 mg (1.4 mmol) of 2 in 3.5 ml of absolute ethanol under an argon atmosphere. Hydrogen was slowly bubbled through the stirred mixture. The reaction was monitored by TLC (Rf 0.30;n-butanol/H2O/HOAc, 25:5:3, v/v). After filtration, solvents were evaporated to yield 3 as a colorless oil (352 mg, 83%). 352 mg (1.2 mmol) of 3 and 108 mg (1.2 mmol) of glyoxylic acid × H2O were dissolved in 6.8 ml of absolute MeOH. After the addition of 242 mg (3.5 mmol) of NaCNBH3, the solution was stirred for 24 h at RT. Solvents were evaporated, and the resulting residue was subjected to column chromatography (silica gel; MeOH/EtOAc, 3:7, v/v). Fractions with material having a Rf value of 0.44 (TLC, MeOH) were combined, and solvents were evaporated. The residue was again subjected to column chromatography (Lobar RP-18; MeOH/H2O, 3:7, v/v). Fractions with target material (above TLC) were combined, and solvents were evaporated to yield 4 as a colorless oil (101 mg, 24%). Electrospray ionization-mass spectrometry (ESI-MS):m/z 361 [M + H]+; 1H NMR (CD3OD): δ (ppm) 1.35–1.83 (m, 6 H), 1.43 (s, 9 H), 1.45 (s, 9 H), 2.99 (t, 2 H, J = 8.0 Hz), 3.49 (s, 2 H), 3.93 (dd, 1 H, J = 5.0, 9.0 Hz);13C NMR (CD3OD): δ (ppm) 24.00, 26.72, 28.29, 28.75, 32.10, 48.36, 50.67, 55.59, 80.46, 82.60, 158.08, 171.11, 173.52. 182 mg (0.51 mmol) of 4 was dissolved in 9.1 ml of a triethylamine solution (10% in absolute MeOH, v/v). 627 mg (2.75 mmol) of di-t-butyl-dicarbonate was added, and the reaction mixture was refluxed for 30 min while stirring. Stirring was continued for 30 min at RT, the solvents were evaporated, and the resulting residue was taken up in ice-cold 1 n HCl. The emulsion was immediately extracted with EtOAc. The organic layer was dried with MgSO4, solvents were evaporated, and the residue was subjected to column chromatography (Lobar RP-18; MeOH/H2O, 7:3, v/v). Fractions with material having a Rf value of 0.57 (TLC; MeOH/EtOAc, 3:7, v/v) were combined, and solvents were evaporated to yield 5 as a colorless oil (202 mg, 86%). ESI-MS: m/z 461 [M + H]+; two diastereoisomers were identified by NMR:1H NMR (CD3OD): δ (ppm) 1.26–1.80 (m, 6 H), 1.42/1.43/1.45/1.46 (s/s/s/s, 27 H), 3.27 (t, 2 H,J = 7.0 Hz), 3.87/3.91 (s, 2 H), 3.93 (m, 1 H);13C NMR (CD3OD): δ (ppm) 23.94/24.12, 28.31, 28.60, 28.77, 28.92, 32.37, 49.31/49.85, 50.05, 55.73, 80.33, 81.37, 82.40, 157.28/157.52, 157.96, 173.32/173.49, 173.70. To a solution of 202 mg (0.43 mmol) of5 in 4.8 ml of dry tetrahydrofurane, 70 mg (0.43 mmol) of 1,1-carbonyldiimidazole was added. The reaction mixture was stirred 3 h at RT, and 130 mg (0.43 mmol) of 3 was added. Stirring was continued for 24 h. Solvents were evaporated, and the resulting residue was subjected to column chromatography (silica gel; EtOAc/n-hexane, 1:1, v/v). Fractions with material having a Rf value of 0.26 (TLC, same solvents) were combined, and solvents were evaporated to yield6 as a colorless oil (126 mg, 39%). ESI-MS:m/z 745 [M + H]+; two diastereoisomers were identified by NMR: 1H NMR (CD3OD): δ (ppm) 1.27–1.80 (m, 12 H), 1.43 (s, 18 H), 1.46 (s, 27 H), 3.19 (t, 2 H, J = 7.0 Hz), 3.27 (t, 2 H, J = 7.0 Hz), 3.78/3.82 (s/s, 2 H), 3.92 (dd, 1 H, J = 5.2, 9.0 Hz); 13C NMR (CD3OD): δ (ppm) 24.28, 28.31, 28.72, 28.77, 30.06, 32.35, 32.41, 40.08, 49.05/49.27, 49.39/49.85, 55.74, 80.34, 81.46, 82.43, 157.52, 158.03, 171.76, 173.71. 80 mg (0.11 mmol) of 6 was dissolved in a mixture of 1 ml of tetrahydrofurane and 1 ml of 6 n HCl. The reaction was monitored by TLC (Rf 0.14;n-butanol/H2O/HOAc/pyridine, 4:2:3:3, v/v). Solvents were completely removed, and the residue was dried over KOH under high vacuum to yield GOLA as a colorless amorphic material (38 mg; 78%; GOLA × 3HCl salt based on elemental analysis). Accurate mass (mean of 10 measurements ± S.D.): m/z333.2144 ± 0.0013 [M + H]+ (333.2138, calculated for C14H29N4O5);1H NMR (CD3OD): δ (ppm) 1.42–2.07 (m, 12 H), 3.07 (t, 2 H, J = 8.6 Hz), 3.27 (t, 2 H,J = 7.0 Hz), 3.82 (s, 2 H), 3.97 (t, 1 H,J = 6.2 Hz), 4.00 (t, 1 H, J = 6.4 Hz);13C NMR (CD3OD): δ (ppm) 23.18, 23.31, 26.50, 29.67, 30.90, 31.03, 39.92, 48.40, 49.42, 53.60, 53.84, 166.37, 171.70, 171.82. 13 μl of concentrated HCl was added to a suspension of 1.0 g (13.2 mmol) of glycolic acid in 3.3 g (39.3 mmol) of 3,4-dihydro-2 H-pyrane. The reaction mixture was heated to 80 °C for 30 min. Volatiles were evaporated, and the resulting residue was taken up in hexane and subjected to column chromatography (silica gel; EtOAc/n-hexane, 3:7, v/v). Fractions with material having a Rf value of 0.18 (TLC, same solvents; charring with H2SO4) were combined, and solvents were evaporated to yield 7 as a colorless oil (0.50 g, 24%). Chemical ionization-MS:m/z 233 [M + H]+ after trimethylsilylation. Because 7 was unstable, it was immediately used for synthesis of 8 without further characterization. 160 mg (1.0 mmol) of 7 and 162 mg (1.0 mmol) of 1,1-carbonyldiimidazole were dissolved in 3.8 ml of dry tetrahydrofurane and stirred for 3 h at RT. 302 mg (1.0 mmol) of3 was added, and stirring continued for 24 h at RT. Solvents were then evaporated, and the resulting residue was subjected to column chromatography (silica gel; EtOAc/n-hexane, 1:1, v/v). Fractions with material having a Rf value of 0.18 (TLC, same solvent) were combined, and solvents were evaporated to yield 8 as a colorless oil (207 mg, 47%). ESI-MS:m/z 445 [M + H]+; 1H NMR (CD3OD): δ (ppm) 1.35–1.95 (m, 12 H), 1.43 (s, 9 H), 1.45 (s, 9 H), 3.24 (t, 2 H, J = 7.0 Hz), 3.52 (ddd, 1 H, J = 1.4, 4.0, 11.0 Hz), 3.84 (ddd, 1 H, J = 3.4, 7.6, 11.0 Hz), 3.92 (dd, 1 H,J = 5.2, 8.6 Hz), 3.97 (d, 1 H, J = 15.6 Hz), 4.11 (d, 1 H, J = 15.6), 4.63 (dd, 1 H,J = 2.6, 4.4 Hz); 13C NMR (CD3OD): δ (ppm) 20.42, 24.15, 26.33, 28.30, 28.76, 30.04, 31.32, 32.36, 39.61, 55.70, 63.67, 67.54, 80.29, 82.37, 100.61, 157.95, 172.32, 173.62. 180 mg (0.41 mmol) of 8 was dissolved in a mixture of 2.5 ml of tetrahydrofurane and 2.5 ml of 3 nHCl. The reaction was monitored by TLC (Rf 0.48, same as for GOLA). Solvents were completely removed, and the residue was taken up in H2O and passed over a small chromatography column (RP-18, 40–63 μm; H2O) to remove brown reaction side products. Fractions with target material were combined, solvents were removed, and the resulting residue was dried over KOH under high vacuum to yield GALA as a colorless amorphic material (68 mg, 69%; GALA × 1HCl salt based on elemental analysis). Accurate mass (mean of 10 measurements ± S.D.): m/z205.1188 ± 0.0009 [M + H]+ (205.1188, calculated for C8H17N2O4);1H NMR (CD3OD): δ (ppm) 1.47 (m, 2 H), 1.59 (m, 2 H), 1.93 (m, 2 H), 3.28 (t, 2 H, J = 7.0 Hz), 3.96 (t, 1 H, J = 6.4 Hz), 3.98 (s, 2 H);13C NMR (CD3OD): δ (ppm) 23.21, 29.75, 31.01, 39.69, 53.79, 62.14, 171.69, 175.76. CMPM was synthesized according to Lustenberger et al. (18Lustenberger N. Lange H.-W. Hempel K. Angew. Chem. Int. Ed. Engl. 1972; 84: 255-257Crossref Google Scholar) with major modifications. A solution of 879 mg (4.32 mmol) of pyridoxal in 2 ml of H2O was adjusted to pH 9.3 and then added to a solution of 389 mg (5.18 mmol) of glycine in 2 ml of phosphate buffer (0.5 m, pH 9.3). After 1 h of stirring at 4 °C, 3 ml of a NaBH4 solution (100 mg/ml 0.1 n NaOH) was added, and the mixture was stirred for 1 h at RT. The pH was adjusted to 6.0 with HCl, and the solvents were evaporated. The residue was subjected to column chromatography (Lobar RP-18, H2O). Material with a Rfvalue of 0.45 (TLC; n-butanol/H2O/pyridine/HOAc, 4:2:3:3, v/v) was combined, and solvents were evaporated to yield CMPM as colorless crystals (899 mg, 92%). ESI-MS: m/z227 [M + 1]+; 1H NMR (D2O):δ (ppm) 2.15 (s, 3 H), 3.33 (s, 2 H), 4.08 (s, 2 H), 4.39 (s, 2 H), 7.34 (s, 1 H); 13C NMR (D2O):δ (ppm) 17.19, 44.41, 48.65, 59.39, 127.16, 129.81, 133.71, 148.33, 160.97, 172.68. 197 mg (0.8 mmol) of N α-t-Boc-lysine was dissolved in 4 ml of H2O, and 232 μl (2.0 mmol) of glyoxal solution (40%) and 160 μl (2.0 mmol) of formaldehyde solution (37%) were added. The reaction mixture was heated for 24 h at 50 °C. After cooling, 4 ml of 6 n HCl were added, and the solution was stirred for 30 min at RT. Solvents were then completely removed in a vacuum concentrator, and the resulting residue was subjected to column chromatography (Lobar RP-18; heptafluorobutyric acid (HFBA) eluents, 35% solvent B as described below). Fractions with material having a Rf value of 0.13 (TLC;n-butanol/H2O/pyridine/HOAc, 4:2:3:3, v/v) were combined and immediately subjected to ion exchange chromatography (Dowex 50WX4–400, H+ form). The column was first flushed with water, and the target material was then eluted with 3n HCl. Solvents were evaporated, and the residue was freeze-dried to yield GOLD as a colorless amorphic material (56 mg, 31%; GOLD × 3HCl based on elemental analysis; spectroscopic data compliant with the literature (19Wells-Knecht K.J. Brinkmann E. Baynes J.W. J. Org. Chem. 1995; 60: 6246-6247Crossref Scopus (94) Google Scholar)). In general, incubations were conducted in 0.2 mphosphate buffer, pH 7.4, after sterile filtration in a shaker incubator (New Brunswick Scientific, Nürtingen, Germany) at 37 °C. Reactant concentrations are given in the respective legends to figures and tables. Deaerated conditions were achieved in the presence of 1 mm diethylenetriaminepentaacetic acid and by gassing with argon. The workup for reaction mixtures with N α-t-Boc-lysine was as follows. 25 μl of a NaBH4 solution (84.6 mg/ml 0.01 nNaOH) were added to 250-μl aliquots of incubations. After 1 h at RT, 2 ml of 3 n HCl were added. Solvents were removed after 30 min in a vacuum concentrator (Savant-Speed-Vac Plus SC 110 A combined with a Vapor Trap RVT 400, Life Sciences International, Frankfurt, Germany). The residue was taken up in H2O to appropriate concentrations for final high-performance liquid chromatography (HPLC) analysis. For analysis of CMPM, incubations were directly diluted without reduction and acid treatment. The workup for protein incubations was as follows. Incubations were dialyzed extensively against phosphate-buffered saline (pH 7.4) and then subjected to acid or enzymatic hydrolysis before HPLC analysis. The protein content after dialysis was measured by Coomassie blue staining using a Bio-Rad protein assay kit with BSA as standard. 3-mg aliquots of protein were dried, dissolved in 2 ml of 6n HCl, and heated for 20 h at 110 °C under argon. Volatiles were removed in a vacuum concentrator, and the residue was diluted with H2O to concentrations appropriate for HPLC analysis. 3-mg aliquots of protein were diluted to 1 ml with phosphate-buffered saline and treated with the following enzymes: proteinase K (Sigma catalog no. P6556), carboxypeptidase Y (Sigma catalog no. C3888), peptidase (Sigma catalog no. P7500), Pronase E (Sigma catalog no. P5147), and aminopeptidase (Sigma catalog no. L6007). The enzymes were applied stepwise, and incubations with each were for 24 h at 37 °C in a shaker incubator. Specific combinations and concentrations are given in the legend to Table III. In general, a small crystal of thymol was added with the first digestion step. Once the total digestion procedure was completed, reaction mixtures were filtered through molecular weight cut-off 5000 filters (Roth, Karlsruhe, Germany) and diluted with H2O to concentrations appropriate for HPLC analysis. Efficiencies of acid and enzymatic hydrolyses were compared by the ninhydrin method and HPLC analysis of GOLD.Table IIIEfficiency of enzymatic protein hydrolysis for cross-link structuresMethodsEnzymatic hydrolysis protocolsaCombinations used for 3 mg of highly modified protein: A, 0.2 units of proteinase K/0.6 units of Pronase E/0.3 units of aminopeptidase; B, 0.6 units of Pronase E/0.3 units of aminopeptidase/0.2 units of carboxypeptidase; C, 0.1 units of peptidase/0.6 units of Pronase E/0.3 units of aminopeptidase; D, 0.6 units of Pronase E/0.6 units of Pronase E/2 units of aminopeptidase; E, 0.6 units of Pronase E/0.6 units of Pronase E/2 units of aminopeptidase/1.9 units of carboxypeptidase Y.Acid hydrolysisABCDE%GOLD3944636062100Ninhydrin76721077392100a Combinations used for 3 mg of highly modified protein: A, 0.2 units of proteinase K/0.6 units of Pronase E/0.3 units of aminopeptidase; B, 0.6 units of Pronase E/0.3 units of aminopeptidase/0.2 units of carboxypeptidase; C, 0.1 units of peptidase/0.6 units of Pronase E/0.3 units of aminopeptidase; D, 0.6 units of Pronase E/0.6 units of Pronase E/2 units of aminopeptidase; E, 0.6 units of Pronase E/0.6 units of Pronase E/2 units of aminopeptidase/1.9 units of carboxypeptidase Y. Open table in a new tab A Jasco (Groß-Umstadt, Germany) ternary gradient unit 980-PU-ND with degasser, autosampler 851-AS, column oven CO-200 set at 25 °C, and fluorescence detector 920-FP was used. Chromatographic separations were performed on stainless steel columns (VYDAC 218TP54, 250 × 4.6 mm, RP18, 5 μm, Hesperia; Knauer Eurospher 100, 250 × 4.6 mm, RP18, 5 μm, Knauer, Berlin, Germany) using a flow rate of 1.0 ml/min. For the detection of CML, GOLA, and GOLD, the VYDAC column was used with the mobile phase consisting of water (solvent A) and MeOH/water (7:3, v/v; solvent B). 1.2 ml/liter HFBA was added to both solvents A and B. A postcolumn derivatization reagent was added at 0.5 ml/min prior to the detector. This reagent consisted of 0.8 g of o-phthaldialdehyde, 24.7 g of boric acid, 2 ml of 2-mercaptoethanol, and 1 g of Brij 35 in 1 liter of H2O adjusted to pH 9.8 with KOH. The effluent was monitored at 340 nm (excitation)/455 nm (emission). The gradient for CML was as follows: 0–15 min, 2% solvent B isocratic; 15–20 min, linear gradient to 100% solvent B; hold for 10 min. The gradient for GOLA and GOLD was as follows: 0–115 min, 6% solvent B isocratic; 115–120 min, linear gradient to 100% solvent B; hold for 10 min. For CMPM, the VYDAC column was used with the above HFBA solvents but without postcolumn derivatization. The effluent was monitored at 328 nm (excitation)/395 nm (emission). The gradient for CMPM was as follows: 0–15 min, 5% solvent B isocratic; 15–25 min, linear gradient to 30% solvent B; 25–30 min, linear gradient to 100% solvent B; hold for 10 min. For GALA, a 2-stage HPLC system with final precolumn derivatization was used. First, material was collected from the VYDAC column with the above HFBA solvents but without postcolumn derivatization (tR 6.6–9.3 min). The gradient was the same as for CML. Solvents were evaporated in a vacuum concentrator, and to the resulting residue 10 μl of the above postcolumn reagent were added and vigorously mixed. After 2 min 190 μl of phosphate buffer (50 mm, pH 6.6) were added, and the solution was mixed and immediately subjected to the second HPLC system. The Knauer column was used with the mobile phase consisting of H2O/MeOH (9:1, v/v, 50 mm phosphate buffer, pH 7.0; solvent A) and H2O/MeOH (2:8, v/v; solvent B). The gradient was as follows: 0–45 min, 22% solvent B isocratic; 45–50 min, linear gradient to 100% solvent B; hold for 10 min. The effluent was monitored at 340 nm (excitation)/455 nm (emission). For the detection of GOLD the postcolumn derivatization system described above was used. For the detection of GOLA a two-stage HPLC system was used. Material was first collected from the VYDAC column with the above HFBA solvents and gradient (tR 69–86 min) but without postcolumn derivatization. Solvents were evaporated in a vacuum concentrator, and the resulting residue was reconstituted in H2O and subjected to the second HPLC system using the above postcolumn derivatization. The Knauer column was used with the mobile phase consisting of phosphate buffer (1.13 g of NaH2PO4 × 2H2O/liter, pH 4.0, solvent A) and propanol/water (6:4, v/v, 1.0 g of NaH2PO4 × 2H2O/liter, pH 4.0, solvent B). To both solvents A and B, 3 g of sodium dodecylsulfate/liter were added. The gradient was as follows: 0–50 min, linear gradient 20–40% solvent B; 50–55 min, linear gradient to 100% solvent B; hold for 10 min. Water-insoluble protein was obtained from the laboratory of R. H. Nagaraj (prepared according to Ref. 20Padayatti P.S. Ng A.S. Uchida K. Glomb M.A. Nagaraj R.H. Investig. Ophthalmol. Vis. Sci. 2001; 42: 1299-1304PubMed Google Scholar) and enzymatically hydrolyzed following protocol C (TableIII). For detection of GOLA, a two-stage HPLC system was used. First, material was repeatedly collected from the VYDAC column with the above HFBA solvent and gradient (tR 69–86 min) without postcolumn derivatization. Solvents of combined fractions were evaporated in a vacuum concentrator, and the resulting residue was then reconstituted in H2O and subjected to the second HPLC system using the above precolumn derivatization. The Knauer column was used with the same mobile phase as for GALA. The gradient was as follows: 0–60 min, 43% solvent B isocratic; 60–65 min, linear gradient to 100% solvent B; hold for 10 min. For low resolution ESI-MS and coupled HPLC/ESI-MS, an HP-MS engine (Agilent Technologies, Hamburg, Germany) connected to the above Jasco HPLC system was used. Target material was first enriched by repeated collection from the above HFBA/VYDAC HPLC system and was then reinjected on a Knauer Eurospher 100 column (100 × 2.0 mm, RP18, 5 μm; the above HFBA eluents; flow rate: 0.15 ml/min; gradient: 0–30 min, linear gradient from 5 to 30% solvent B; 30–35 min, linear gradient to 100% solvent B; hold for 10 min). For accurate mass determination of amino acid modifications (21Tyler A.N. Clayton E. Green B.N. Anal. Chem. 1996; 68: 3561-3569Crossref PubMed Scopus (67) Google Scholar), a Micromass (Manchester, UK) VG platform II quadrupole mass spectrometer equipped with an ESI interface was used (ESI+; source temperature, 80 °C; capillary, 3.0 kV; cone voltage, 25 V). The data were collected in the multichannel acquisition mode with 128 channels per m/z unit using 13 scans (5 s) with a 0.2-s reset time. The resolution was 650 (10% valley definition). The sample was dissolved for analysis in water/MeCN (1:1) containing polyethylene glycol 200 (4 μg/ml) as a reference material, ammonium formate (0.1%), and formic acid (0.5%); the sample concentration was similar to that of polyethylen
Referência(s)