A major pathophysiological consequence of hyperglycae mia is the extensive chemical interaction of glucose with proteins, leading to its attachment to these proteins with out the aid of enzymes. Even though the Maillard reac tions have been of considerable interest to food chemists since the tum of the century,' it has only been relatively recently that attention has focused on non-enzymic glyca tion of proteins in vivo. Although non-enzymic glycation leading to formation of reversible Amadori products acts on many proteins throughout the body, it is less obvious how these products are related to the pathophysiology of diabetic complications. Recent efforts have focused on biologically important further products of the glycation reaction, which are derived slowly from the Amadori product following a sequence of further reactions and rearrangements? These compounds, in contrast to the Amadori product, are formed irreversibly resulting in accumulation on long-lived proteins; these have been called advanced glycation end (AGE) products.
BIOCHEMISTRY OF THE EARLY MAILLARD REACTIONThe initial Maillard reaction is the condensation of the free aldehyde group of carbohydrate with either the f-amino group of lysine or hydroxylysine residues or the a-amino group of the N-terminal amino acid of proteins.2 Only open forms of sugars react with proteins, the carbonyl group of an acyclic monosaccharide attaching to a protein amino group via nucleophilic attack to form a labile aldi mine (Schiff base).2 This product may hydrolyse back to glucose and protein or undergo an Amadori rearrange ment to form a l-amino-I-deoxyfructose (fructosamine) derivative by a stable, though slightly reversible, keto amine linkage (Fig. 1). This product can cyclise to a ring structure (N-substituted-I-amino-deoxyketopyranose ).2The rate of the Amadori rearrangement is approximately one-sixtieth that of the dissociation to glucose and protein3 and also varies between proteins: for example, the Ama- dori rearrangement occurs about five times more rapidly in albumin than in haemoglobin.4 The labile Schiff base form can also exist as a cyclic glucose adduct (N-substituted aldosylamine) and in vivo proteins will exist predom inantly in the cyclic form of both the Schiff base and the Amadori product, although the labile form is probably lost in most purification procedures.4Glycation of haemoglobin is somewhat atypical, as this reaction occurs predominantly between glucose and the N-terminal valine of the �-chain of haemoglobin to form HbA,/ on other proteins glucose adducts are found on lysine and hydroxylysine residues.2 A variety of other aldose and ketose sugars, including glucose-6-phosphate, galactose, mannose, ribose, fructose and xylulose, can participate in the glycation reaction.6 The relative reactiv ity varies up to 300-fold between monosaccharides and depends largely on the equilibrium between the reactive open (carbonyl) and closed (hemiacetal) configurations of the sugar. Aldose sugars react more rapidly than ketose sugars since the aldeh...