This work describes the demonstration of a kilogram-scale synthesis of a γ-secretase modulator from a plantsterol-derived starting material. Key to developing a synthetic route capable of delivering a kilogram of the target compound was the development of a four-reaction telescope process and a selective O-alkylation of a triol intermediate. These improvements enabled the successful delivery of kilogram-scale batches of API for preclinical development studies.
■ INTRODUCTIONDuring the course of research efforts focused on developing a γ-secretase modulator (GSM) as a potential treatment for Alzheimer's disease (AD), we discovered a new structural class of GSMs. 1 A recent series of publications documents our early work on this series of molecules, which is derived from a complex plant triterpene found in black cohosh (Actaea racemosa). 2−4 Semisynthetic derivatives featuring a C3 morpholine acetal and a C24 ether on the modified triterpene scaffold emerged as particularly potent GSMs. 2,3 Further refinement of the series led to the discovery of SPI-1865 (1), which was ultimately selected as a candidate for further development. 5 In order to provide 1 in the quantities required for our preclinical and early clinical development efforts, significant improvements to the synthetic route were required to enhance the overall efficiency and throughput. In this publication, we discuss the synthetic challenges associated with the synthesis of 1 and describe how we were able to develop a multireaction telescope sequence and selective O-alkylation conditions that allowed us to execute the kilogram-scale synthesis of the target compound.
■ RESULTS AND DISCUSSIONFirst-Generation Synthesis. The initial medicinal chemistry route to 1, designed to install a variety of substituents on the azetidine nitrogen, utilized our natural-product-derived starting material 2 obtained from the crude root extract of black cohosh. 6 The synthesis commenced with a double oxidative cleavage of C3 terpene glycosides 2 (Scheme 1). 7,8 The resulting dialdehyde could be used without purification in the subsequent double reductive amination with 1-Boc-3-aminoazetidine hydrochloride to provide N-Boc-azetidinylmorpholine 3. Reduction of the C15 ketone proceeded with high diastereoselectivity (>95:5 dr) to give the C15 alcohol, and purification provided 4 in 61% yield over three steps when corrected for the purity of starting material 2. 9 In order to avoid any potential problems with the selectivity of a downstream Oalkylation, the C15 hydroxyl group was protected as the triethylsilyl ether prior to hydrolysis of the C24 acetate to unveil the C24,C25 diol 5. We found that we could selectively alkylate the C24 hydroxyl over the tertiary C25 hydroxyl to provide C24 ethyl ether 6 (73% yield over two steps). Finally, the synthesis of the target compound was completed by employing a two-step process involving initial double deprotection of the N-Boc-carbamate and the C15 triethylsilyl ether under acidic conditions followed by reductive amination with aceton...