The design and synthesis of stereodiverse scaffolds

This thesis describes our efforts to use stereochemistry as a diversity element in the design and synthesis of aspartyl protease inhibitors. Three different acyclic scaffolds are described, as well as related derivatives accessed through further diversification of the scaffold backbone. These scaffolds were embedded into peptides representing the recognition sequence of two enzymes, renin and β-secretase. They were inserted over the cleavage site where it was envisioned they would mimic the developing tetrahedral character of the transition state, providing tight binding inhibitors.; Peptides containing scaffold A were screened for their ability to inhibit renin. These peptides were moderate inhibitors with the most potent compounds having IC₅₀ values in the low micromolar range. Interestingly, these peptides displayed very different lipophilicities among the various stereoisomers, inspiring us to consider the use of stereochemical diversity not only to optimize ligand affinity, but perhaps to modulate other pharmacological parameters. Peptides containing scaffold B were synthesized in an effort to improve the potency of renin inhibitors. This second library resulted in several low nanomolar inhibitors, indicating that scaffold B is more suited to inhibitors of renin than scaffold A. Importantly, a stereochemical preference among these inhibitors was discovered which would not have been predicted based on literature precedent.; As an extension of our results with renin, we synthesized β-secretase peptides containing scaffold B and assayed them for inhibition. Quite surprisingly, these peptides were only modest inhibitors, displaying IC₅₀ values in the 100–200 μM range. Application of scaffold C, which is related to scaffold B by a methylene transposition, provided similar levels of moderate inhibition. The loss of potency was postulated to derive from the more open orientation of the active site cleft of β-secretase compared to renin, making it a more challenging target due to the decreased surface area available for attractive interactions. Comparison of the results between these two superficially similar enzymes highlights the need for diversity of all types in the identification of biologic ligands and underscores the need for unbiased approaches to ligand discovery.