Progress in the development of beta-secretase inhibitors for Alzheimer's disease.

Abstract

Since the original identification of BACE in 1999 and until quite recently, BACE was often regarded as a "difficult" drug target, much as renin has proven to be. The reasons for this include the following. First, the long and shallow nature of the substrate binding pocket suggested that it would not be possible to identify small molecule drugs that could have adequate binding affinity. Second, functional groups that typically interact with the active site aspartates are usually highly polarized and, therefore, contribute to reduced CNS localization. Early BACE inhibitors were all designed using knowledge of the peptide substrates and usually contained some variation of a few well-known transition-state isosteres. While these had great impact on fundamental understanding of the enzyme structure and key interaction regions, they were very large, very polar, and had essentially no CNS availability. Continued progress by reducing the peptidic nature of these compounds resulted in incremental advances and has provided compounds that meet, or nearly meet, typical CNS drug-like criteria. The challenges associated with peptidic starting points inspired innovative new approaches to search for different starting points. Several groups employed high concentration screening (ligand concentration 100 microM and higher) to find weak hits after conventional screening (typically at 10 microM) failed to find more potent ones. Fragment-based methods have also been developed to identify even weaker hits (IC50 1 mM and greater). This was accomplished through the evolution and refinement of several detection methodologies including calorimetry, surface plasmon resonance, NMR, and crystallography. Coupled with detailed structural understanding of ligand-enzyme interactions and focus on maintaining ligand efficiency, these developments have resulted in several examples where potency was improved by 10,000-fold to afford compounds with IC50 values < 10 nM and promising drug-like characteristics. Together, all these efforts have afforded a diverse array of chemotypes as BACE inhibitors. Early work focused on improving BACE potency in isolated enzyme assays. However, most of these compounds showed potency reductions in cellular assays. Continued improvements in drug properties and in understanding of the physiologically relevant conditions have resulted in many compounds that show strong potency in both isolated and cellular assays. Several compounds have shown reduction of Abeta using rodent in-vivo models both peripherally and in the brain. Recently, one compound has demonstrated reduction of brain Abeta levels in a non-human primate. Phase I clinical trials were initiated on BACE inhibitor CTS-21166 from CoMentis in July of 2007. This compound derives from the earliest described peptidic inhibitors such as OM99-2 [58] but no details have been reported. In addition to strategies involving small molecule inhibitors of BACE and gamma-secretase to reduce Abeta levels, the application of biological agents has been under investigation since the identification of Abeta. The earliest efforts in this area failed. Despite encouraging results in preclinical models, immunization against Abeta by administration of AN-1792 from Elan led to development of aseptic meningoencephalitis in 6% of the patients receiving the drug. Nevertheless, continued efforts with other biological approaches appear encouraging. Most advanced in clinical trials is bapineuzumab from Elan, which is in Phase III clinical trials. This is a humanized monoclonal antibody against Abeta plaques. A recent monograph is devoted to progress in these areas. Taken together, considerable progress has been made in developing CNS-penetrant agents that reduce AP levels and in providing validation that such agents will be therapeutically beneficial for the treatment of Alzheimer's disease.