Pergamon 0031-9422(95)00211-1 Phytochemistry, Vol. 40, No. 3, pp. 773 784, 1995 Copyright © 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved 0031 9422/95 $9.50 + 0.00 BIOSYNTHESIS OF IRIDOIDS IN S YRINGA AND FRAXINUS: SECOIRIDOID PRECURSORS* SOREN DAMTOFT, HENRIK FRANZYK and SOREN ROSENDALJENSEN Department of Organic Chemistry, The Technical University of Denmark, DK-2800 Lyngby, Denmark (Received 19 December 1994) Key Word Index--Fraxinus excelsior; Syringa josikaea; Syringa vulgaris; Oleaceae; secoiridoid gluco- sides; biosynthesis; deuterium labelling; oleosides; oleuropein; taxonomy. Abstract--Several deuterium-labelled secoiridoids have been prepared and tested as possible precursors for the iridoids in Fraxinus excelsior, Syringa josikaea and S. vulgaris. Oleoside 11-methyl ester was an efficient precursor for the oleosides, whereas secologanin-type iridoids gave only significant incorporation in S. josikaea. In this plant low incorporations into the oleosides were also seen for kingiside and 8-epi-kingiside. The major pathway to the oleosides therefore seems to proceed via a direct ring fission of ketologanin to oleoside 11-methyl ester. A Baeyer-Villiger-like mechanism which explains the different compounds found in the plants is proposed, and the taxonomy of the Oleaceae is discussed. Due to the unique presence of the usual pathway leading to secologanin and its congeners in Fontanesia, the Oleaceae is considered to be a member of the Gentiananae rather than Scrophulariales/Lamianae. INTRODUCTION In the preceding papers [I,2] we have shown that the biosynthesis of oleosides (i.e. secoiridoids with an 8,9- double bond) proceeds via 7-ketologanic acid (I) or ketologanin (2), and that it thus differs from the biosynth- esis of the secologanin-type iridoids, where neither I nor 2 are intermediates. Possible secoiridoid precursors for the oleosides were considered some time ago and two routes were discussed I-3]. The first route proceeds via secologanin (3), which might undergo rearrangement of the 8,10-double bond to compound 4 and subsequent oxidation at C-7 to oleoside l l-methyl ester (5). The second route was through 8-epi-kingiside (6), which by anti-elimination could give 5. To discriminate between these alternatives, feedings of [II-O14CH3]-3, [II- O14CH3]-6 and [I l-Ol4CH3]kingiside (7) to Olea euro- paea were carried out. Both lactone epimers (6 and 7) gave the same incorporations (0.13 %) into oleuropein ($), whereas secologanin (3) gave a slightly higher incorpora- tion (0.34%). The two kingisides (6 and 7) gave even lower incorporations (0.02%) into jasminine (9) in Jas- minum primulinum [3,4]. We have noted that iridoid- producing plants seem to be rather effective in converting externally supplied precursors [1, 2 and refs therein], and therefore significant results with late-stage iridoid precur- sors require at least 1% incorporation. Consequently, we * Part 3 in the series 'Biosynthesis of iridoids in Syringa and Fraxinus' for part 2, see ref. [2]. consider all these reported incorporations to be of doubt- ful significance and a pathway through the kingisides thus seems unlikely. The next study [5] was based on the assumption that secologanin (3) was a common inter- mediate in the biosynthesis of the oleosides as well as of the ligustalosides (secoiridoids with a 10-aldehyde func- tionality, e.g. 10 and 11). Assuming that the next step is epoxidation of the 8,10-vinylic bond of 3, the oleosides might be formed by a reductive opening of the epoxide (8S-12) with subsequent elimination of water, and the formation of 10-hydroxy-oleosides and ligustalosides could thus be rationalized. Nevertheless, when feeding the epoxy-secologanins 8R-12 and 8S-12 as well as both epimers of epoxy-secoxyloganin (8S-13 and 8R-13) to three oleaceous plants: Olea europaea, Osmanthusfraorans and Ligustrum japonicum, the largest incorporation into the oleosides was 0.14% while 8S-12 gave 0.51% incor- poration into 10. Again, the incorporations were not convincing, and as a continuation of our work with carbocyclic precursors [1,2] we decided to reinvestigate potential secoiridoid precursors for the oleosides in plants with a known, high biosynthetic capacity. In this work we have thus synthesized deuterated ana- logues of 8-epi-kingisidic acid (14) and kingisidic acid (15) as well as the methyl esters 6 and 7. Additionally, sec- ologanin (3), secoxyloganin (16) and secologanoside (17) have been prepared in labelled form together with oleo- side l 1-methyl ester (5) and the oleosidic secologanin analogue 4. They have been tested as precursors for the iridoids found in Fraxinus and Syringa. These com- pounds were discussed in the preceding paper [2]. 773