br Discussion br Author Contributions br

Discussion
Author Contributions
Alzheimer’s disease (AD) is an age-related, chronically progressive neurodegenerative disorder affecting more than 35million people worldwide and an estimated 5.5million in the US. Current marketed drugs such as acetylcholinesterase inhibitors and the NMDA antagonist memantine only treat disease symptoms and do not slow or reverse the underlying progression of the disease. Thus, the development of disease-modifying treatments remains a major interest in the pharmaceutical industry. Pathological examination of AD brain tissue has revealed two underlying disease components, neurofibrillary tangles and amyloid plaques. Plaques are formed from the precipitation of the amyloid beta peptides (Aβ), which range from 37 to 42 amino acids in length and are produced from sequential cleavages of the amyloid precursor protein (APP) by β-secretase 1 (BACE 1) and γ-secretase. While the exact cause of the disease is still an active area of research, a current prevailing hypothesis suggests that oligomers of the 42-amino 97 9 form of the beta amyloid peptide (Aβ1–42) are central to the disease process. Efforts to reduce Aβ1–42 production by interfering with the action of γ-secretase have led to the identification of full inhibitors of the enzyme complex (γ-secretase inhibitors, or GSIs), several of which have been advanced into clinical trials. This approach, however, is complicated by issues including inhibition of Notch processing by unselective GSIs and the clinical failure of the GSIs semagacestat and avagacestat. These issues have led to the continued development of additional treatment options. A recent additional approach is to target the γ-secretase complex with molecules that change the length of the Aβ peptides produced. This class of molecules, known as γ-secretase modulators (GSMs), are believed to change the processing of APP-CTFβ (the substrate for γ-secretase) so that shorter, more soluble peptides (such as Aβ1–38) are produced instead of the highly insoluble and neurotoxic Aβ1–42. Because they do not block γ-secretase processing, GSMs do not interfere with Notch signaling, and it is possible that they may avoid some of the issues of nonselective GSIs. The first generation of GSMs, including NSAID derivatives such as ibuprofen, indomethacin and tarenflurbil generally have weak in vitro potency (Aβ1–42 IC=25–200μM) and were not effective in the clinic. Second generation GSMs have been disclosed by a number of companies and have demonstrated efficacy in a number of preclinical AD animal models. Based on the interesting biological profile of the GSM class, we screened the BMS compound collection for compounds that inhibit Aβ1–42 production in a cellular assay. Active compounds were then rescreened for their effect on total Aβ production to identify potential GSMs. To our delight, a class of diaminotriazines represented by compound was identified as having modest potency for inhibition of Aβ1–42 production (IC=180nM) with no effect on total Aβ production (). Our initial efforts to optimize the diaminotriazine scaffold have already been described, including the observation that lengthening the triazine aryl substituent to a benzyl substituent and changing the triazole group to a methyl imidazole as in compound resulted in a nearly 10-fold increase in potency. We were also able to incorporate heterocycles other than triazine as the core of the molecule. In an effort to increase potency by conformational restriction, we questioned whether an additional ring could be constructed linking the aniline ring back to the triazine or other heterocycle. The resulting compounds would be somewhat related to the well-known tricyclic antidepressants, and might be expected to have reasonable brain penetration (). We selected a thiazole ring as the first heterocycle simply for the synthetic ease of using the well-established Hantszch cyclization to construct the ring. The synthesis of these initial analogs is outlined in .