hide
Free keywords:
-
Abstract:
Based on the initial studies in the Fürstner group, the herein reported synthetic efforts mark the first total synthesis of putative chagosensine. After proving the mis-assignment of the natural product by the isolation team, we embarked on the synthetic endeavor for structural elucidation of the natural product by synthesizing seven additional diastereoisomers.
Embedded in a 16-membered macrocycle, the intriguing polyketide possesses two trans-configured THF-rings and a unique Z,Z-configured 2-chloro-1,3-diene. In addition to the structural confirmation of the complex molecular architecture inhabiting a total of 11 stereogenic centers, a concise and scalable synthesis was planned in order to provide sufficient quantities of the natural product as well as analogues for biological profiling.
In the initial period of the synthetic endeavor we focused on furnishing the 16-membered macrocyclic core. Hence, neither a ring-closing alkene nor alkyne metathesis was able to forge the strained ring in sufficient quantities, which prompted us to pursue a macrolactonization approach. A novel palladium-catalyzed cross coupling assembled the two fragments in a highly convergent and regioselective manner. Subsequent chloro-destannylation completed the first stereoselective introduction of the most salient feature of chagosensine. After liberating the seco-acid, the ring closure of the strained 16-membered macrocycle was accomplished by lactonization under forcing Mukaiyama conditions for the first time. Despite the successful synthesis of the core cycle of chagosensine, the synthetic route suffered from a low overall yield in combination with a putative unselective introduction of the missing side chain, which led to a revised synthetic route.
By maintaining the established key transformations, the implementation of the entire northern fragment was envisioned to enable the first total synthesis of this fascinating natural product. Despite the successful synthesis of the complete northern fragment based on the previous fragment synthesis the trans-hydrometalation followed by protodemetalation raised unsurmountable difficulties. Therefore the revised, second-generation synthesis of the complete northern part commenced with (−)-citronellal as the ultimate surrogate for the side chain, finally affording the northern fragment in a remarkable overall yield of 7.6% over 15 steps.
Subjecting the complete northern fragment to the optimized reaction conditions for the site-selective Stille cross coupling and subsequent chloro-destannylation, the unique Z,Z-chlorodiene was introduced in high yield as a single isomer. Undesirable differentiation of a diol-containing seco-acid resulted in the exclusive formation of the highly strained 13-membered macrocycle. However, protection group management enabled the ring closure to the desired 16-membered macrolactone in 40% yield under forcing Yamaguchi lactonization conditions. Global deprotection and subsequent Pinnick oxidation liberated the putative natural product, which surprisingly proved to be highly unstable. Derivatization to the methyl ester afforded a stable compound, enabling comparison with the derivatized natural product. Unfortunately, this methyl ester mismatched with the isolated methyl ester, displaying major discrepancies in 1H- and 13C-NMR along the entire carbon skeleton. After unambiguously proving the stereochemical mis-assignment by the isolation team, the final stage of this thesis aimed for the elucidation of the correct stereochemistry of chagosensine.
After identifying structural subunits with questionable stereochemical assignment, the stereodivergent synthesis of seven putative diastereomers was planned by maintaining the elaborated late-stage functionalizations. In regard of the second northern fragment, the previous route was adjusted in a single transformation paving the way for installing all other stereogenic centers. In contrast to the modular synthesis of the isomeric northern fragment, the scalable synthesis of the four southern fragments required a new route.
Based on the newly developed cross coupling protocol, a compound library of eight diastereoisomers was generated from the two northern fragments in combination with the four southern fragments. Following the reliable sequence, the synthesis of the additional seven seco-acids was accomplished by parallel synthesis. During the ring closure, the isolated yield ranged from 11% to 72% due to inevitable substrate dependency. Although the final synthetic protocol enabled the isolation of all stereoisomers, the putative natural products were highly fragile. Treatment with the methanolic NMR solvent mixture either induced ring expansion by intramolecular transesterification to the stable 18-membered macrocycle or opened the macrolactone at ambient temperature. Neither the data sets of the stable derivatives nor the 1H-NMR spectra of the mixtures with unstable diastereomers are matching with the literature data. In addition to the tremendous NMR-spectroscopic deviations, further deep-seated discrepancies with the isolation paper could be revealed, which in combination with the unavailability of the original spectra make the structural revision of chagosensine an impossible task.
