Abstract

The reaction of lens proteins with sugars over time results in the formation of protein-bound advanced glycation end products (AGEs). The most damaging element of AGE formation may be the synthesis of protein-protein cross-links in long-lived proteins, such as collagen or lens crystallins. A quantitative cross-linking assay, involving the sugar-dependent incorporation of [U-(14)C]lysine into protein, was employed to determine the efficacy of a variety of potential cross-linking inhibitors. Reaction mixtures contained 5.0 mM L-threose, 2.5 microCi [(14)C]lysine (1.0 mCi/mmole), 5.0 mg/ml bovine lens proteins, 0-10 mM inhibitor and 1.0 mM DTPA in 100 mM phosphate buffer, pH 7.0. Of 17 potential inhibitors tested, 11 showed 50% inhibition or less at 10 mM. The dicarbonyl-reactive compounds 2-aminoguanidine, semicarbazide and o-phenylenediamine inhibited 50% at 2.0 mM, whereas 10 mM dimethylguanidine had no effect. Several amino acids failed to compete effectively with [(14)C]lysine in the cross-linking assay; however, cysteine inhibited 50% at 1.0 mM. This was likely due to the sulfhydryl group of cysteine, because 3-mercaptopropionic acid and reduced glutathione exhibited similar activity. Sodium metabisulfite had the highest activity, inhibiting 50% at only 0.1-0.2 mM. Protein dimer formation, as determined by SDS-PAGE, was inhibited in a quantitatively similar manner. The dicarbonyl-reactive inhibitors and the sulfur-containing compounds produced similar inhibition curves for [(14)C]lysine incorporation over a 3 week assay with 250 mM glucose. A much lesser effect was observed on either the incorporation of [(14)C]glucose, or on fluorophore formation (360/420 nm), suggesting that non-cross-link fluorophores were also formed. The inhibitor data were consistent with cross-linking by a dicarbonyl intermediate. This was supported by the fact that the inhibitors were uniformly less effective when the 5.0 mM threose was replaced by either 3.0 mM 3-deoxythreosone or 3.0 mM threosone.