A. IntroductionWhen intestinal sucrase and pancreatic a-amylase were identified at Bayer as new targets for improved diabetes therapy (PULS et al. 1973), the problem remained of the best way to find a potent and selective inhibitor of these enzymes. The medicinal chemist today has two alternatives in the search for a first lead compound, firstly the screening of thousands of compounds with maximum structural diversity in a random screening approach, or secondly the rational design of a lead compound, when the biochemical mechanism and the structure of the enzyme are weIl known. In the late 1960s, when we started to look for such potent and selective inhibitors, we had to choose the random screening approach, relying mainly on testing of extracts of the culture broths of microorganisms.Today the mechanism of enzymatic splitting of sucrose by intestinal sucrase is fairly weIl understood. First the glucosidic oxygen atom of sucrose is protonated via a carboxyl group of the enzyme. The splitting of the glucosidic C-O bond is further facilitated by the glucosyl cation being stabilized by a second carboxylate group of the enzyme (COGOLI and SEMENZA 1975). FinaIly, the glucosyl cation reacts with water to give the products 0-glucose and o-fructose. The protonated sucrose molecule land the glucosyl cation 11 are two high-energy intermediates of the enzymatic reaction. Today we know that mimics of such high-energy intermediates are often potent inhibitors of the enzyme. Accordingly, compounds similar in structure to o-glucose but with an easily protonated basic N-atom either in the position of the anomeric oxygen atom (inhibitor type I) or in the position of the ring oxygen atom (inhibitor type 11) should be potent inhibitors of sucrase (Fig. 1). Interestingly, both types of inhibitors represented by either acarbose or I-deoxynojirimycin were found in our random screening. In the foIlowing sections these two types of inhibitors are not discussed in a historical order but under structural criteria. Not included in this discussion are the a-amylase inhibitors of protein nature because there is no evidence that they will find any therapeutic application.