| dc.description |
[2⠲-<sup>14</sup>C]-7-Hydroxy-2⠲,4⠲,5⠲-trimethoxyisoflavone (2b) is shown to be excellently converted into [6-<sup>14</sup>C]-amorphigenin (1b) by Amorpha fruticosa seedlings, this being the first direct demonstration of the biosynthetic link between isoflavones and rotenoids. Isoflavones rather than isoflavanones are implicated and prenylation is evidently a post-rotenoid phase. In agreement, (±)-[6-<sup>3</sup>H]-9-demethylmunduserone (8a) is a good precursor for amorphigenin. This places 9-demethylmunduserone in a key biosynthetic position as parent of a sub-family of rotenoids, similar unprenylated rotenoids heading other sub-families. (±)-[6-<sup>3</sup>H]-8-Dimethylallyl-9-demethyl-munduserone and [6-<sup>3</sup>H]rotenone (as 6aS,12aS,5⠲R- and 6aR,12aS,5⠲S-diastereoisomers) are each satisfactorily incorporated into amorphigenin. Post-rotenoid stages are thus likely to be dimethylallylation, epoxidation, cyclisation to dalpanol (17), dehydration to rotenone (1a), and 8⠲-oxidation to amorphigenin. Administration of chalcones shows that the 2⠲,4,4⠲-trihydroxy-member is the most acceptable precursor and a mechanism for 2,3-aryl migration in biosynthesis takes account of the free 4-hydroxy-group requirement. The product from the rearrangement appears to be 7-hydroxy-4⠲-methoxyisoflavone rather than the 4⠲,7-dihydroxy-compound. Methylation accompanying rearrangement of a spirodienone intermediate (14) can accommodate this. Acetate is incorporated into rings D and E, but the radioactivity from serine and glycine is found in risic acid (16) obtained by degrading amorphigenin. This is discussed. |
