著者

樋口 隆一 時光 義徳 古森 徹哉

出版者

天然有機化合物討論会実行委員会

雑誌

天然有機化合物討論会講演要旨集

巻号頁・発行日

vol.28, pp.224-231, 1986

---

Mechanisum of the diazomethane degradation for the sugar-aglycone linkage of the gypsogenin 3-O-glycoside(1) was studied and the useful degradative reaction was applied to the structure determination of quillayasaponin. Since three kinds of functional groups of 1, the 4α-CHO group in the aglycone, COOH and 4-OH group in the glucuronic acid were presumed to contribute to the degradative reaction, the necessarity of the each functional group was examined using a model compound 7, and only the 4α-CHO group was showed to be essential. By taking the above evidence and the usual reaction products of aldehyde with diazomethane into account, the reaction must be proceeded through such a oxide intermediate as shown in the scheme 6. QS-III(25), a major acylated genuine bisdesmoside of the so-called quillayasaponin, afforded a 28-O-glycosidal triterpenoid(29) still possessing acyl moiety by this diazometane degradation. On the basis of chemical and spectral evidence, especially, comparison of the ^<13>C NMR spectra of 29 and its desacyl compound(30)(Table 1), the site of linkage of acyl moiety of 29 was determined. The structure of 25, therefore, was complex acylated oligosaccharide as shown in scheme 8.

著者

滝川 浩郷 森 謙治 木戸 勝 Albizati K. F. Faulkner D. J.

出版者

天然有機化合物討論会実行委員会

雑誌

天然有機化合物討論会講演要旨集

巻号頁・発行日

vol.34, pp.707-714, 1992

---

In 1985, limatulone (1a and 1b) was isolated from the intertidal limpet Collisella limatula, and found to inhibit fish and crab predation. It is the most potent fish feeding inhibitor and is about an order of magnitude more effective than polygodial. Although the natural limatulone was optically inactive, it was not clear whether it was a meso-compound (1a) or a racemate (1b). We therefore decided to confirm the structure by a total synthesis. The known 3, which was obtained from the starting material (2) in 4 steps, was alkylated with BrCH_2CO_2Et to give 4a. The corresponding acid 4b was employed for the lactone-ring formation to give a stereoisomeric mixture of lactones 5a and 5b. After separation, lactone 5a was converted to the coupling-partners 10 and 11 via 6, respectively. The carbanion derived from 11 was alkylated with 10 to give a complex mixture, which was desulfonylated with Na-Hg to give a mixture of 12a and 12b. This was converted to a separable mixture of 13a and 13b. Fortunately, the structure of the less polar isomer could be solved by an X-ray analysis. An aldehyde 15a, which was prepared from 13a in 3steps, was treated with ClCH_2Li to give bis-epoxide 16a. Epoxide-opening of 16a by a Grignard reagent gave 17a. Swern oxidation of 17a to 18a was followed by removal of the EE protective groups to give meso-limatulone (1a), (overall yield: 0.62% from 2 in 24 steps). Similarly, (±)-13b was converted to (±)-limatulone (1b), (overall yield: 0.39% from 2 in 24 steps). The ^1H-NMR spectrum of the reported limatulone was identical to that of (±)-limatulone (1b). To our surprise, however, another fraction from the HPLC of Collisella limatula showed the ^1H-NMR spectrum superimposable on that of meso-limatulone (1a). Accordingly, Collisella limatula produces both meso-1a and (±)-1b.