Peridinin isomer

To link the two half moieties of the molecule, a Julia-Kocienski olefmation was carried out between the C19 building block 59 (again prepared by syn-SN2 -substitu-tion of a propargylic oxirane with DIBAH) and the C20 building block 60, formed via oxidation of 58 with Mn02 (Scheme 18.19). Although this reaction initially led to the formation of the Z-isomer as the major product, the latter was readily isomerized at room temperature to the desired all-trans-polyene peridinin (6). 

Katsumura and collaborators [137-139] reported the total synthesis of a polyfimctional carotenoid, Peridinin 314 (Scheme 101). Diene-allenyl sulfone 312 was combined with unsaturated aldehyde 313 using NaHMDS in THF. The product was obtained in 50% yield as a mixture of isomers EjZ 25/75. 

The synthon (117) is derived from the optically active C22-allenic apocarotenal 118 which is prepared via a Wittig condensation of the (3S)-Ci5-allenic aldehyde 103 with the Cy-phosphonium chloride 119. Reduction of the aldehyde group in 118 (a mixture of 11Z and 11 ) with NaBH4, followed by acetylation, yields the acetate 120, which is converted into the sulphone 117 by heating under reflux with sodium sulphinate in propan-2-ol and water. Condensation between the (all- )-allenic sulphone 117 and the formyl ester (5R,6S)-116 under the conditions of the sulphone method furnishes a mixture (ca. 1 1) of the optically active peridinin (108) and its (11 )-isomer and these are cleanly separated by preparative HPLC in the dark Scheme 27) [70,72]. 

The key step of the synthesis is the rearrangement of the a-acetylenic alcohol 97 into the unsaturated carbonyl compound 124. This rearrangement was carried out with tris(triphenylsilyl)vanadate, triphenylsilanol and benzoic acid to give a mixture of the isomers 124 and 125. The latter was converted by iodine catalysis into the desired isomer 124. This key intermediate was afterwards transformed into the Cis-phosphonium salt 123 by standard procedures. The Wittig olefination of the Cio-dial 45 first with the fucoxanthin end group 123 and then with the peridinin end group 122 gave, in five steps, the C4o-carotenoid 126. Finally the epoxidation of this compound resulted in optically active fucoxanthin (121) and its (5S,6R)-isomer (Scheme 28). 

H.p.l.c. separation of the (—)-camphanic diesters of astaxanthin [3,3 -dihy-droxy-jS,/8-carotene-4,4 -dione (11)] from lobster eggs (Homarus gammarus) showed that all three isomers, (3/2,3 /2), (35,3 5), and (3/2,3 5), were present. This is the first identification of a carotenoid in a natural source." Details of the previously reported" determination of the chirality of peridinin [(35,5/2,6/2,3 5,5 /2,6 5)-5, 6 -epoxy-3,5,3 -trihydroxy-6,7-didehydro-5,6,5, 6 -tetrahydro-10,ll,20-trinor-/8,/8-caroten-19,ll -olide 3-acetate (17)] and dinoxanthin [(35,5/2,6/2,3 5,5 /2,6 5)-5, 6 -epoxy-6,7-didehydro-5,6,5, 6 -tetrahydro-/3,/5-carotene-3,5,3 -triol 3-acetate (12)] have been given."" Sul-catoxanthin from Anemonia sulcata has been shown to be identical to peridinin." ... 

As shown above in the synthesis of compounds 21 and 22, allenic synthons are produced by DIBAH reduction of 5,6-epoxy-7,8-ethynyl structures (Scheme 4). Many steps are needed, however, to prepare the chiral epoxyethynyl esters 19 and 20 via 16 and 17 from the simple starting material 2. On the other hand [15-17], the chiral key intermediate 21b can be prepared by analogous DIBAH reduction of the epoxyethynyl diacetate 31 which is obtained, together with the (Zj-isomer by the treatment of optically active 4 with MCPBA (Scheme 7). The latter method has been applied to the synthesis of the allenic part of peridinin, shown in Section D. 

Condensation between the (all- ) allenic sulphone 51 and the formyl ester 45 under the conditions of the sulphone method furnishes a mixture (ca. 1 1) of the optically active peridinin (558) and its (1 l )-isomer and these are cleanly separated by preparative HPLC in the dark [17,21] (Scheme 10). 

NMR was first used in structural elucidation of carotenoids for peridinin (30) (108, 160, 161). Support for the presence of the allo-xanthin (31) end group in isomytiloxanthin (20) was obtained by direct comparison of NMR spectra (114, 175). From the characteristic shifts of the 14-methyl and C-8 signals violeoxanthin was shown to be the 9-ds isomer of violaxanthin (32) (136). 

---