A simple, quantitative assay has been established to determine the glycation-dependent crosslinking ability of any sugar by measuring the incorporation of (\left[{ }^{14} \mathrm{C}\right]) lysine into protein. The assay was shown to be both sugar-dependent and protein-dependent and was completely inhibited by sodium cyanoborohydride, 2-aminoguanidine, and semicarbazide. A typical (1.0-\mathrm{ml}) reaction mixture contained (5 \mathrm{mg}) lysozyme, (20 \mu) mol threose, and (5 \mu \mathrm{Ci}) of (\left[{ }^{14} \mathrm{C}\right]) lysine and exhibited an incorporation of (8 \times 10^{5} \mathrm{cpm}) ( (1.6 \mathrm{nmol}) of lysine) after 7 days of incubation. A comparison of the crosslinking ability of a variety of sugars showed glyceraldehyde and dihydroxyacetone to be twice as active as erythrose and threose and eight times more reactive than ribose. Little or no crosslinking could be demonstrated with three different hexoses as well as their phosphorylated derivatives. The dicarbonyl sugars 3-deoxyglucosone and xylosone were at least as effective as ribose in crosslinking, as were the oxidation products of ascorbic acid. Several amine-containing compounds were tested as inhibitors of crosslinking; however, 2 -aminoguanidine was the most effective. The rate of synthesis of Lys-Lys, LysArg, and Lys-His crosslinks was determined by measuring the incorporation of (\left[{ }^{14} \mathrm{C}\right]) lysine into specific amino acid homopolymers. The relative incorporation was polylysine (>) polyarginine (>) polyhistidine with threose, but polyarginine (>) polyhistidine (>) polylysine with dehydroascorbic acid, suggesting a different crosslinking mechanism for these two compounds. (c. 1994 Academic Press, Inc.