Total Synthesis of Fidaxomicin and Analogs

<p>Fidaxomicin (1), also known as Tiacumicin B or Lipiarmycin A3, belongs to the tiacumicin class of antibiotic natural products and has recently been approved by the FDA and EMA for the clinical treatment of Clostridium difficile infections. In addition, this antibiotic has also been found to be effective against multi-drug resistant Mycobacterium tuberculosis (< 0.1 mg/L). Fidaxomicin and other members of the tiacumicin family are characterized by a synthetically interesting, highly unsaturated macrolide with multiple stereogenic centers as well as two carbohydrates attached to the core moiety by β-glycosidic bonds. Therefore, the synthetic challenge lies in the rapid construction of each fragment, concise assembly of these building blocks in a stereoselective manner as well as the selection of suitable protecting groups.</p><p>Recently, we accomplished the first enantioselective total synthesis of fidaxomicin in a highly convergent manner. Herein, we will present our investigations leading to the synthesis of fidaxomicin and analogs with the aim to establish structure activity relationships.</p><p>The synthesis commenced with the preparation of resorcylate 6 applying a biomimetic aromatization cascade reaction (Scheme 2). The rhamnose intermediate 9 was synthesized from α-methyl-D-mannopyranoside in 6 steps (Scheme 3). The key esterification with rhamnoside part 9 and resorcylate 6 via a ketene intermediate afforded the desired ester 10 with excellent regioselectivity for the O-4 position. After several functional group transformations, rhamnose donor 11was obtained. Next, we opted to prepare noviosyl donor 17 for the challenging noviosylation on the sterically hindered alcohol 33 (Scheme 4). We selected a cyclic carbonate for protection of the 2,3-diol with the aim to achieve high β-selectivity in the noviosylation due to the carbonate’s electron-withdrawing characteristics.</p><p>We then turned our attention to the synthesis of the macrocyclic building block starting from the known alcohol 23 and s-hydroxy ester 18to obtain boronic ester 25 for the Suzuki coupling (Scheme 5). The alkenyl halide for Suzuki coupling was synthesized applying a Vinylogous Mukaiyama Aldol Reaction (VMAR) as a key step (Scheme 6). After optimization, this transformation was performed at -37 °C with careful addition of N,O-acetal 27 to give the desired anti-alcohol 29 in 75% yield. Following straightforward functional group transformations gave access to the secondary alcohol 33 for noviosylation. </p><p>Noviosylation of 33 turned out to be very challenging, but the use of HgO/HgBr₂ gave good β selectivity (α:β = 1:3) and satisfactory yield (Scheme 7). Even in the presence of the sensitive carbonate protecting group, the following Suzuki coupling gave the macrocycle precursor 35 in excellent yield. Subsequent Ring Closing Metathesis (RCM) furnished the macrocycle 36 after selective deprotection of the silyl group. The β-selective rhamnosylation was achieved in good yield after extensive screening of reaction conditions. Finally, careful deprotections and purification by reversed-phase HPLC gave pure synthetic fidaxomicin (1). The identity of the synthetic compound was confirmed by co-injection of the isolated, authentic natural product on reversed-phase HPLC, and by ¹H NMR analysis. Interestingly, some of the chemical shifts of the natural product and the synthetic compound did not completely match each other. Therefore, we mixed equimolar amounts of the fully synthetic product and isolated fidaxomicin, and thus confirmed the identity of the two compounds by ¹H NMR spectroscopy.</p>