Superoxide free radical generation by Amadori compounds: the role of acyclic forms and metal ions.

Generation of oxygen free radicals by glycated proteins is widely believed to be one of the causes of oxidative stress in diabetes and aging. Metal ion catalysis is regarded as an essential part of the oxidative mechanism. In this work, we also considered an alternative "metal-free" superoxide radical formation by a number of fructose-amino acids (Amadori compounds) derived from glycine and lysine, which represent the simplest models for early glycated proteins. In the superoxide dismutase-dependent cytochrome c assay, 1 mM Chelex-treated aqueous solutions of monofructose-amino acids 4-6 generated 0.9-3.6 x 10(-10) M s-1 O2*- at pH 7. Surprisingly, the rates of superoxide radical formation in the solutions of difructose-amino acids 7-9 were significantly higher (0.75-5.8 x 10(-9) M s-1 O2*-). The percentage of acyclic sugar anomers (

Inhibition of colony formation in agarose of metastatic human breast carcinoma and melanoma cells by synthetic glycoamine analogs.

We studied the influence of 10 synthetic glycoamine analogs on colony formation in 0.3 and 0.9% agarose by metastatic human breast carcinoma (MDA-MB-435) and melanoma (TXM-13) cells. Nine synthetic analogs significantly inhibited the colony formation in 0.9% agarose of MDA-MB-435 human breast carcinoma cells; five compounds caused a 73-83% reduction of colony formation. Seven synthetic glycoamines caused a significant inhibition of colony formation in 0.9% agarose by TXM-13 melanoma cells with the inhibitory effect ranging from 71 to 87%. The 50% inhibition (I50) doses and relative activity rank of the compounds were similar for both breast carcinoma and melanoma cell lines. The murine B16 melanoma cell aggregation assay was employed to elucidate the potential mechanism(s) of the inhibitory activity of synthetic glycoamines. The relative activity ranks of the compounds based on the independently determined I50 doses for both cell aggregation and clonogenic growth assays were very similar for the four most active synthetic analogs and clearly indicated the importance of hydrophobic amino acid in mediating the bioactivity of synthetic glycoamines. In both experimental systems (clonogenic growth in agarose and cell aggregation assay) the leading compound was N-(1-deoxy-D-fructos-1-yl)-D-leucine (Fru-D-Leu) and the least active analog was N-(l-deoxy-D-fructos-1-yl)-glycine (Fru-Gly). These results show that synthetic glycoamines may act by competing for specific carbohydrate-lectin interactions, particularly those involving beta-galactoside-specific lectins expressed on metastatic cells.

Molecular and crystal structure of N-(2-deoxy-D-aldohexos-2-yl)-glycines (Heyns compounds).

Heyns compounds, 2-carboxymethylamino-2-deoxy-D-glucose (1), -mannose (2), and -galactose (3), were prepared by N-carboxymethylation of the corresponding hexosamines and 1 was also prepared via the reaction of D-fructose with glycine. Both 1 and 3 crystallize from aqueous solutions as zwitterions in the alpha-pyranose form and in the 4C1 conformation. Crystalline 1 is nearly isostructural to N-acetylglucosamine, forming stacks of molecules with infinite chains of homodromic hydrogen bonds along the stacks. For both 1 and 3, all hydroxyl, ammonium, and carboxyl groups are involved in intermolecular hydrogen-bonding, and an intramolecular hydrogen bond in 3 is formed via interaction of the ammonium and carboxyl groups. 1H and 13C NMR spectra (D2O solutions) indicate that all of the compounds are conformationally unstable, and that the major form present in D2O solution at 25 degrees C is the 4C1 alpha-pyranose form, with the 4C1 beta-pyranose form present in lesser amounts. In addition, for solutions of 2 and 3, considerable amounts of alpha- and beta-furanose forms are present and exist in conformations favorable for a cis-relationship between the carboxymethylammonium and anomeric hydroxyl groups.

Catalysis of reduction of carbohydrate 2-oxoaldehydes (osones) by mammalian aldose reductase and aldehyde reductase.

In mammalian tissues, carbohydrate 2-oxoaldehydes, or 'osones', formed by cleavage of carbohydrate residues from glycated proteins, cause damage to cells and tissues by cross-linking of proteins. In the substrate specificity study reported here, we show that several osones are relatively good substrates for the reduced, unactivated form of aldose reductase (EC 1.1.1.21) from human and pig muscle, and aldehyde reductase (EC 1.1.1.2) from pig kidney, enzymes that have been well characterised both structurally and mechanistically. Since these enzymes are relatively ubiquitous, they may serve to protect a large number of tissues from damage, by catalysing the reduction of locally-produced osones. Reduction of all substrates by aldehyde reductase obeyed Michaelis-Menten kinetics. In contrast, a Hill constant of about 0.5 was obtained for aldose reductase-catalysed reduction of each of the carbohydrate 2-oxoaldehydes, and for several other substrates that were examined. Although this deviation from Michaelis-Menten kinetics has been ascribed to the presence of two forms of the enzyme, activated and unactivated, our results suggest that it is a characteristic of the unactivated form.

The reaction of D-glucose with aminoguanidine.

The reaction of D-glucose with aminoguanidine was examined at pH 7.0 and 37 degrees C (phosphate buffer). Under these conditions, the reaction requires ca. 42 days for 50% of the sugar to react, as measured by the disappearance of D-glucose, and at 60 degrees C all the aminoguanidine had reacted within 72 h. The initial product, a beta-D-glucopyranosyl aminoguanidine (1) was obtained in the crystalline state as the trifluoroacetate salt. Data collected on this compound suggests that, in solution, it is largely a glycosylamine in the beta pyranose form. Acetylation gave a crystalline heptaacetate (2), which, in solution (as evidenced by NMR spectroscopy) exists in two different conformational forms. The crystal structure of the heptaacetate also includes two conformers. Both crystallographically independent molecules are in the normal beta pyranose form, with the acetylated guanyl residue occupying different spatial positions relative to the ring.

Crystal structure of an Amadori compound, N-(1-deoxy-beta-D-fructopyranos-1-yl)-glycine ("D-fructose-glycine").

The first crystal structure data on an Amadori compound, N-(1-deoxy-beta-D-fructopyranos-1-yl)-glycine, are reported. The space group is P2(1) with Z = 2 and cell parameters a = 7.246(1), b = 10.009(1), c = 7.060(1) A, and beta = 101.085(6) degrees. The structure was solved by direct methods and refined to a final R of 2.9% and Rw of 3.8% for 1385 reflections to give esd's of 0.002 A in bond lengths and 0.2 degree in angles. The conformation of the carbohydrate is the normal 2C5 pyranose chair. Bond lengths and valence angles compare well with average values from a number of pyranose structures. The molecule of the Amadori compound exists in the zwitterion form and has the C-6-O-6-C-2-C-1-N-C-2'-C-1'-O-1' chain in a zig-zag conformation, that is (together with O-2') substantionally planar. All hydroxyl, carboxyl, and ring oxygen atoms, and the secondary ammonium group are involved in hydrogen bonding, which forms a three-dimensional network of two infinite chains that have an ammonium group as a common segment. The shortest intra- and inter-molecular hydrogen bonds involved donors of the pyranosyl moiety and acceptors of the amino acid portion, and vice versa.

3-Deoxyfructose concentrations are increased in human plasma and urine in diabetes.

3-Deoxyglucosone (3-DG) is a reactive dicarbonyl sugar thought to be a key intermediate in the nonenzymatic polymerization and browning of proteins by glucose. 3-DG may be formed in vivo from fructose, fructose 3-phosphate, or Amadori adducts to protein, such as N epsilon-fructoselysine (FL), all of which are known to be elevated in body fluids or tissues in diabetes. Modification of proteins by 3-DG formed in vivo is thought to be limited by enzymatic reduction of 3-DG to less reactive species, such as 3-deoxyfructose (3-DF). In this study, we have measured 3-DF, as a metabolic fingerprint of 3-DG, in plasma and urine from a group of diabetic patients and control subjects. Plasma and urinary 3-DF concentrations were significantly increased in the diabetic compared with the control population (0.853 +/- 0.189 vs. 0.494 +/- 0.072 microM, P < 0.001, and 69.9 +/- 44.2 vs. 38.7 +/- 16.1 nmol/mg creatinine, P < 0.001, respectively). Plasma and urinary 3-DF concentrations correlated strongly with one another, with HbA1c (P < 0.005 in all cases), and with urinary FL (P < 0.02 and P = 0.005, respectively). The overall increase in 3-DF concentrations in plasma and urine in diabetes and their correlation with other indexes of glycemic control suggest that increased amounts of 3-DG are formed in the body during hyperglycemia in diabetes and then metabolized to 3-DF. These observations are consistent with a role for increased formation of the dicarbonyl sugar 3-DG in the accelerated browning of tissue proteins in diabetes.

The preparation and characterization of some Amadori compounds (1-amino-1-deoxy-D-fructose derivatives) derived from a series of aliphatic omega-amino acids.

Amadori compounds (1-amino-1-deoxy-D-fructose derivatives) were prepared by reacting D-glucose with a series of aliphatic amino acids. These include Amadori compounds derived from glycine (1), beta-alanine (2), gamma-amino butyric acid (3), delta-aminovaleric acid (4), epsilon-amino-caproic acid (5) and N alpha-formyl-L-lysine (6). In the FAB mass spectra, molecular-ion clusters as well as fragment ions corresponding to loss of water or CO2 molecules were observed. The 13C NMR spectra indicate that all the compounds are conformationally unstable, but that the predominant form present in solution (D2O) is the beta-pyranose form. The 1H NMR spectra of 1 and 2 indicate a slow rotation around the C-1-C-2 bond, possibly as a result of an intramolecular hydrogen bond involving the carboxyl group. The pK alpha's of all compounds were measured by pH-potentiometric titration in 0.2 M KNO3 solution at 25 degrees C. All compounds showed a decrease in the basicity of their amino groups (in the order of approximately 1.5 of the K alpha value), and 1 and 2 showed a decrease in the basicity of their carboxyl groups (in the order of approximately 0.2) in comparison with that of parent amino acids.

Ascorbic acid glycation: the reactions of L-threose in lens tissue.

L-Threose is a significant degradation product of ascorbic acid at pH 7.0 in the presence of oxygen. When compared to several other ascorbate-derived degradation products, it had the greatest ability to glycate and crosslink lens proteins in vitro. To determine whether L-threose was formed in the lens, the sugars in a TCA-soluble extract from human lenses were reduced to polyols with NaBH4, acetylated and analysed by gas-liquid chromatography. The threitol levels measured were 3.4 +/- 0.8 micrograms per lens (n = 4). GC-MS measurements made after reduction with NaBD4 indicated that threitol, but little or no threose, was originally present in the human lens. Rat lenses were incubated with [1-13C]D-threose for 24 hr, and considerable D-threitol formation was seen by NMR spectroscopy. Analysis of the lenses after medium removal showed that only [1-13C]threitol was present within the lenses indicating a rapid reduction of threose within the lens, presumably by aldose reductase. Assays with human recombinant aldose reductase and with human lens cortical and nuclear extracts all exhibited sorbinil-inhibitable aldose reductase activity with L-threose as substrate. This was confirmed by incubating a preparation of [1-14C]L-tetrose (a mixture of 40% L-threose and 45% L-erythrose) with both the pure aldose reductase and crude lens extracts followed by the subsequent identification of the [1-14C]L-threitol formed by thin layer chromatography. L-Threose degrades very slowly in 0.1 M phosphate buffer at pH 7.0, but the addition of a four-fold excess of N alpha-acetyl-L-lysine accelerated the rate of disappearance of threose 30-fold, indicating a rapid glycation reaction. When [1-14C]L-tetrose was incubated with a complete bovine lens homogenate, a linear incorporation into protein was observed over a 24 hr period. Increasing levels of lens extract exhibited increasing incorporation into protein. These data confirm a rapid reactivity of L-threose with lens protein and argue that glycation would occur in vivo in spite of the presence of aldose reductase.

The degradation of L-threose at Maillard reaction conditions.

L-Threose, a comparatively unstable aldose, is produced from L-ascorbic acid in the presence of oxygen and participates vigorously in Maillard reactions, even at comparatively mild conditions. In the present study, the degradation of L-threose at pH 7.0 alone, in the presence of N-alpha-acetyl-L-lysine, and at pH 2.0 alone at 37 degrees C was investigated by identification of some of the products produced in the reactions by means of GLC and GLC-MS. Among the compounds identified were 3-deoxy-tetros-2-ulose (1), the predicted alkaline rearrangement product derived from 1 (2,4-dihydroxybutyrate, the 4-carbon metasaccharinic acid), as well as glyceraldehyde. Isotopic tracer studies clearly show that the glyceraldehyde is produced by loss of C-1 from the starting L-threose molecule. The presence of N-acetyl lysine in incubation solutions appears to accelerate the production of 1, but the formation of glyceraldehyde appears to be independent of the lysine derivative.

The extent of N epsilon-(carboxymethyl)lysine formation in lens proteins and polylysine by the autoxidation products of ascorbic acid.

The autoxidation of ascorbic acid (ASA) leads to the formation of compounds which are capable of glycating and crosslinking proteins in vitro. When the soluble crystallins from bovine lens were incubated with ASA in the presence of sodium cyanoborohydride, a single major adduct was observed, whose appearance correlated with the loss of lysine. When polylysine was reacted with equivalent amounts of ASA under the same conditions, this product represented half of the total lysine content after four weeks of incubation at 37 degrees C. This adduct was isolated and identified as N epsilon-(carboxymethyl)lysine (CML) by TLC, GC/MS and amino acid analysis. Several oxidation products of ASA were each reacted with polylysine in the presence of sodium cyanoborohydride to identify the reactive species. CML was the major adduct formed with either ASA and dehydroascorbic acid (DHA). Markedly diminished amounts were seen with L-2,3-diketogulonic acid (DKG), and L-threose, while no CML was formed with L-threo-pentos-2-ulose (L-xylosone). In the absence of sodium cyanoborohydride the yield of CML was similar with each of the ASA autoxidation products and required oxygen. Reactions with [1-14C]ASA gave rise to [14C]CML, but only with NaCNBH3 present. At least two routes of CML formation appear to be operating depending upon whether NaCNBH3 is present to reduce the putative Schiff base formed between lysine and DHA.

The reaction of some dicarbonyl sugars with aminoguanidine.

The reactions of aminoguanidine (guanylhydrazine) with 3-deoxy-D-erythro-hexos-2-ulose (1a), 3-deoxy-D-glycero-pentose-2-ulose (1b), D-erythro-hexos-2-ulose (1c), and D-glycero-pentose-2-ulose (1d) were examined at 37 degrees at a solution pH of 7.0 (phosphate buffer). For 1a and 1b, two major products were observed and shown respectively to be the 5- and 6-substituted 3-amino-1,2,4-triazine derivatives. The ratios of the products were independent of the amount of aminoguanidine present or the order of mixing the reagents prior to the experiments. For 1c and 1d, only the 5-substituted triazine derivatives were formed. No evidence for hydrazone or bishydrazone formation was observed.

Detection of 3-deoxyfructose and 3-deoxyglucosone in human urine and plasma: evidence for intermediate stages of the Maillard reaction in vivo.

3-Deoxyglucose (3-deoxy-D-erythro-hexos-2-ulose) (3-DG) is a reactive dicarbonyl intermediate involved in the polymerization and browning of proteins by glucose in vitro. Damage to protein by formation of 3-DG in vivo is thought to be limited by enzymes which convert 3-DG to less reactive species, such as 3-deoxyfructose (3-DF). We have developed a sensitive and specific assay for measuring 3-DG and 3-DF in human urine and plasma. In this assay, 3-DG and 3-DF are reduced to 3-deoxy-hexitols (3-DH), using either NaBH4 or NaBD4, and then analyzed by selected ion monitoring gas chromatography-mass spectrometry. Based on comparative analysis of samples reduced with NaBD4 versus NaBD4, 3-DH in urine was derived exclusively (greater than 99%) from 3-DF, while 3-DG accounted for approximately 15% of 3-DH in plasma. The concentrations of 3-DH in fasting human urine and plasma were 5.3 +/- 1.5 micrograms/mg creatinine (n = 18) and 7.2 +/- 1.7 micrograms/dl (n = 18), respectively. The concentrations of 3-DG and 3-DF in plasma (n = 7) were 1.0 +/- 0.2 and 6.7 +/- 1.6 micrograms/dl, respectively. These results suggest that several milligrams of 3-DG are formed in the body per day and detoxified by reduction to 3-DF and support the role of 3-DG as an intermediate in the browning of protein via the Maillard reaction in vivo.

Glycation of lens proteins by the oxidation products of ascorbic acid.

Bovine lens water-soluble proteins were incubated with [I-14C]ascorbic acid (ASA) for 6 days, and the incorporation into protein was measured at daily intervals. Aliquots were also withdrawn to determine the distribution of label among the various ASA oxidation products. A linear incorporation into protein was observed in the presence of NaCNBH3, however, little or no incorporation was seen in its absence. TLC analysis showed a complete loss of ASA by day 3, whereas both dehydroascorbate (DHA) and diketogulonic acid (DKG) remained constant for 6 days, consistent with the linear incorporation into protein. The amino acid composition of the proteins glycated in the presence of NaCNBH3 was identical to controls except for a 70% reduction in lysine residues and a corresponding increase in an unknown product which eluted slightly earlier than methionine. In the absence of NaCNBH3 lysine decreased linearly to 20% with an additional decrease in arginine and histidine at later times concurrent with protein crosslinking. DHA and DKG were prepared and incubated directly with lens proteins for an 8 day period. Both compounds glycated lens protein as evidenced by an increased binding to a boronate affinity column. SDS-PAGE showed that both compounds were also capable of causing protein crosslinking. DHA is apparently capable of reacting directly with protein since glycation was observed with the ASA analog, reductic acid, which can be oxidized to dehydroreductic acid, but which cannot be hydrolyzed to an open chain structure. DHA also produced a lysine adduct which was not obtained with DKG, supporting the idea that both species have glycating ability.

Detection of Kestoses and Kestose-Related Oligosaccharides in Extracts of Festuca arundinacea, Dactylis glomerata L., and Asparagus officinalis L. Root Cultures and Invertase by C and H Nuclear Magnetic Resonance Spectroscopy.

A previous study (KL Forsythe, MS Feather [1989] Carbohydr Res 185: 315-319) showed that (13)C nuclear magnetic resonance spectroscopy can be used to detect and identify mixtures of 1-kestose and neokestose after conversion to the acetate derivatives. In this study, unequivocal assignments are made for the anomeric carbon and proton signals for the above two trisaccharide acetates as well as for 6-kestose hendecaacetate and for nystose tetradecaacetate (a 1-kestose-derived tetrasaccharide). A number of oligosaccharide fractions were isolated from several plant species, converted to the acetates, and nuclear magnetic resonance spectra obtained. Using the above reference data, the following information was obtained. The trisaccharide fraction from Dactylis glomerata L. stem tissue and Asparagus officinalis L. roots contain both 1-kestose and neokestose, and the tetrasaccharide fractions contain three components, one of which is nystose. Penta- and hexasaccharide acetates were also isolated from A. officinalis L. roots and were found to contain, respectively, four and at least five components. All components of both of the above species appear to contain a kestose residue and to be produced by the sequential addition of fructofuranosyl units to these. The trisaccharide fraction from Festuca arundinacea is complex, and contains at least five different components, two of which appear to be 1-kestose and neokestose.

Modification of galactose and N-acetylgalactosamine residues by oxidation of C-6 hydroxyls to the aldehydes followed by reductive amination: model systems and antifreeze glycoproteins.

Amino acids and peptides have been attached to the C-6 hydroxyls of the galactose and the N-acetylgalactosamine by first oxidizing the C-6 hydroxyls to the aldehydes by galactose oxidase in the presence of small amounts of catalase, followed by reductive amination (alpha-amino group) in the presence of cyanoborohydride. The activity of oxidized antifreeze glycoprotein was greater than 70% of the original, and considerable activity has been retained with some substitutions on reductive amination using cyanoborohydride. The following were some activities retained (as compared with the oxidized antifreeze glycoprotein): Gly, 64; (Gly)2, 88; (Gly)3, 82; (Gly)4, 70; Gly-Gly-NH2, 44; Gly-Glu, 13; Gly-Leu, 40; Gly-Tyr, 57; Gly-Gly-Leu, 50; Gly-Gly-Phe, 30; and Gly-Gly-Val, 35. On amino acid analysis of acid hydrolysates, some release of the amino acid attached by amination occurred; e.g., Gly-Tyr gave 0.26 Gly and 0.49 Tyr per disaccharide.

The precipitation and cross-linking of lens crystallins by ascorbic acid.

Bovine lens beta-crystallin was incubated with increasing concentrations of sugars and sugar derivatives for a period of 2 weeks in the dark at 37 degrees C. Marked protein precipitation and a browning reaction was observed with both ascorbic acid (ASA) and dehydroascorbic acid (DHA), but little or no reaction was seen with several other sugars and sugar analogs. Similar incubations were carried out with 20 mM ASA, 20 mM DHA and 20 mM glucose, but with increasing amounts of the individual crystallins. Glucose was capable of precipitating gamma-crystallin in the presence of air, but this reaction was decreased if dithiothreitol and a chelating agent were added prior to incubation. ASA and DHA produced precipitation and browning with gamma- and beta-crystallin, but not with alpha-crystallin or lens soluble proteins. Similar reactivities were observed both in air and under reducing conditions. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis of these reaction mixtures showed little or no cross-linking with any of the lens proteins by glucose. ASA and DHA caused detectable dimer formation with gamma-crystallin, but produced the formation of dimers as well as highly polymerized proteins at the top of the gel with all the other crystallins and with lens soluble proteins. A time-course experiment with alpha-crystallin in the presence of air showed no cross-linking with 100 mM glucose over a 6-week period; however, 10 mM ASA caused definite cross-linking at 2 weeks, and at 6 weeks a dark smear of protein was visible throughout the gel. ASA was still capable of inducing cross-linking under low oxygen conditions but the protein smearing was markedly diminished. Further, the cross-linking pattern was similar to that seen in the water-insoluble fraction from older human lenses and cataracts. This reaction may be significant in vivo because cross-linking was observed under low-oxygen conditions with as little as 2 mM ASA, which is the level of ASA normally present in human lenses.