Total Synthesis of Tiacumicin A. Total Synthesis, Relay Synthesis, and Degradation Studies of Fidaxomicin (Tiacumicin B, Lipiarmycin A3)

Abstract

The commercial macrolide antibiotic fidaxomicin was synthesized in a highly convergent manner. Salient features of this synthesis include a β-selective noviosylation, a β-selective rhamnosylation, a ring-closing metathesis, a Suzuki coupling, and a vinylogous Mukaiyama aldol reaction. Careful choice of protecting groups and fine-tuning of the glycosylation reactions led to the first total synthesis of fidaxomicin. In addition, a relay synthesis of fidaxomicin was established, which gives access to a conveniently protected intermediate from the natural material for derivatization. The first total synthesis of a related congener, tiacumicin A, is presented.

Total Synthesis of Tiacumicin B: Implementing Hydrogen Bond Directed Acceptor Delivery for Highly Selective β‐Glycosylations

“…b) The rhamnosylation was carried out on the macrolide, using an imidate donor (a/b: 1/4, b: 62 %). Following similar strategies, Gademann also synthesized three congeners of Tcn-B: tiacumicin A [12] and mangrolides A [14] and D. [15] In 2019, de Brabander also reported a total synthesis of mangrolide D. [16] As to our contribution, we reported two related synthetic pathways leading to the aglycone. [17] We designed an original strategy for the synthesis of the C12-C15 diene region of this aglycone and this resulted in the discovery that Pd-nanoparticles catalyze the Kumada-Corriu reaction of vinylsulfides, [18] and also led us to study Griggs allene/alkyne cross-coupling and to propose a mechanism based on DFT calculations.…”

“…b) The rhamnosylation was carried out on the macrolide, using an imidate donor ( α / β : 1/4, β : 62 %). Following similar strategies, Gademann also synthesized three congeners of Tcn‐B: tiacumicin A and mangrolides A and D . In 2019, de Brabander also reported a total synthesis of mangrolide D .…”

“…Reused in this total synthesis, this step should provide a monoglycosylated macrolactone bearing at C11 an unprotected OH directly ready for the β ‐noviosylation step. This scenario was nonetheless uncertain for at least three reasons: a) noviosylation is a very late step, b) a high β ‐selectivity is required, and c) this macrolactone has been described as a reluctant glycosylation acceptor …”

“…Preliminary trials using (+)‐menthol as an acceptor led predominantly to the β ‐adduct ( α / β : 1/5) in a 72 % yield, demonstrating the feasibility of this approach. Unfortunately, the glycosylation of the macrolactonic acceptor 24 was unsuccessful as no adduct was detected . The success of the above‐mentioned rhamnosylation led us to consider that the sulfoxide 35 could be a superiorly reactive donor.…”

“…[7] In 2015 the groups of Gademann [8] and Altmann 9] published the first two syntheses of the aglycone of Tcn-B, and the group of Zhu [10] the synthesis of a diastereomer. Nonetheless, only Gademann was capable of completing the total synthesis of Tcn-B, providing the following answers to the b-glycosylations problem: [11,12] a) The Helferichs protocol (activating agent: HgO (excess) /HgBr 2(cat) ) [13] was used to noviosylate the C4-C13 aglycone fragment (a/b: 1/3, b: 54 %), as the cyclic aglycone gave only a adducts whatever the conditions. b) The rhamnosylation was carried out on the macrolide, using an imidate donor (a/b: 1/4, b: 62 %).…”