NL Swedish

Scheme 3.5.3. APP amino acid sequence close to the cleavage sites and point mutations in A/ annotation.

F45 Stockholm

G45 London a g y

G arctic quired to coordinate and deprotonate water to supply the nucleophilic hydroxyl anion (Scheme 3.5.2), which leads to the tetrahedral intermediate B. Collapse of the hydrated amide releases both a new C-terminal acid and a new N-terminal amine, as in C. Any precursor which gives way to more stable hydrates, e.g. ketones as in D, will interfere with hydrolysis. Electron deficiency of the ketones will stabilize the hydrate even further, therefore a,a-difluoroketones are well established aspartic protease inhibitors. A wide range of isosteric amide replacements was identified from natural protease inhibitors (1, pepstatin, Scheme 3.5.4) or resulted from insightful design. These dipeptide mimetics can be cited in a shorthand version using the three-letter amino acid code to indicate the junction: -C(aaj-aa2)-. The hydroxyethylene isosters E and F were inspired by naturally occurring peptide mimetics, which are known to inhibit aspartic proteases. The specific placement of this pseudo-intermediate between the aspartic acid and the hydrolytic water is reflected by the different stereochemistry of the alcohols in E and F. The neighboring amino acids have to be adapted to the individual requirements of the targeted as-partic protease, be it renin, HIV-protease, plasmepsin, or ¡-secretase.

The high affinity complex of Glu-Val-Asn-C(Leu-Ala)-Ala-Glu-Phe (3, OM99-2) with ¡-secretase results in complete inhibition of ¡-secretase activity and enabled crystallization and structure determination at 1.9A resolution (Figures 3.5.1 and 3.5.2) [22, 23]. The subsite specificity was established by determination of cleavage rates of combinatorial substrate mixtures and resulted in the discovery that Glu-Leu-Asp-C(Leu-Ala)-Val-Glu-Phe was the most potent inhibitor (Ki 0.31 ra) of ¡secretase. Recently a second member of the BACE family with high similarity was identified - BACE2 (also called memapsin 1), which causes additional cleavage reminiscent of a-secretase activity [24].

The pharmacological evidence compiled for y-secretase is indicative of the activity of an aspartic protease requiring at least one additional cofactor. The location of the active site within the membrane makes y-secretase quite unique. Currently, there is only one precedent for a similar, tricky enzyme, signal peptidase, which shares several features and most of the problems associated with inner-membrane location [25]. It will not, unfortunately, be easy to isolate and purify the membrane-stabilized protease while retaining its activity. It has, therefore, so far escaped crystallization and X-ray structure determination. Mutation analysis of the two conserved aspartic acids of all presenilins supports their key role in y-secretase

Scheme 3.5.4.

3.5.2 b-Secretase Inhibitors I 269

activity, however. There is, moreover, evidence for an autoproteolytic mechanism, which is required to deliver active presenilin. The presenilin sheds the exon E9 (Scheme 3.5.7, below) in this cleavage and simultaneous maturation towards active y-secretase. A proposal for the arrangement of the transmembrane helices has been made, but it does not fully explain the observed cleavage pattern [26]. A rudimentary scheme for the eight membrane-spanning domains is depicted below.

0 0

Post a comment