Acetonedicarboxylic Acid Biosynthesis

Scheme 16. New proposal for the biosynthesis of tropane alkaloids via acetonedicarboxylic acid with acetonedicarboxylate followed by loss of only one of the pendant car-boxyl groups. The resulting (22) could then be processed to (19) as shown (Scheme 16).

It is an attractive feature of this proposal that it establishes again a unified conceptual framework for the biosynthesis of tropane and pelletierine type alkaloids.According to this new hypothesis,tropane- and pelletierine-type alkaloids fall into two groups. Group A comprises alkaloids such as N-methylpelle-tierine (23),hygrine (16) and 2,4-dimethylindolizidine (31) in which the carbon atom corresponding to C-3' of (16) or of (23) is not bonded to another atom other than C-2'. These alkaloids would be made by reaction of acetoacetate with the appropriate cyclic imine or iminium ion (Path C1). Alkaloids such as ecgo-nine (19),tropine (1),cuskhygrine (32) or ^-pelletierine (33),on the other hand, belong in group B. These latter alkaloids in which the analogous C-3' carbon atom is linked not only to C-2' but also to another carbon atom, would be formed via a double Mannich condensation of acetonedicarboxylate with (10) or (14). Ketoacids (22) and (29) are postulated as intermediates in the formation of alkaloids belonging to group B.

Me Me O

31 32 33

Me Me O

31 32 33

The experimental data accumulated so far for the origin of the acetate derived C3 chain of (1) are in agreement with this proposal. The presence of two bond-labeling patterns in (1) after administration of [1,2-13C2]acetate has been interpreted hitherto as evidence for the presence of a racemic intermediate. This necessitated the postulation of a nonstereoselective enzymatic oxidation of (22) to (25) as discussed in detail above. However, this postulate in the face of the complete lack of enzymological data in its support, is not entirely satisfying. In contrast, the postulation of acetonedicarboxylic acid, or a derivative thereof, as a pathway intermediate does explain simply the labeling pattern in the C3 side-chain of (1) and a precedent has now been found in the biosynthesis of (27). Moreover,the hypothesis also explains why (21) is not incorporated into tropane alkaloids.

Unfortunately, the new hypothesis does not eliminate all of the apparent confusion surrounding the question of a symmetrical vs a nonsymmetrical intermediate between ornithine and/or arginine and (1) in Datura species. The new model would, however, allow the occurrence of two labeling patterns from [1,2-13C2]acetate in the C3 fragment of (1) to be explained even if nonsymmetrical incorporation of ornithine (4) or arginine (5) is observed. Thus, the apparent contradiction between the results of the early feeding experiments of Leete [13,14,27] and of the Halle group [19] on the one hand and the results from the acetate feeding experiments [22,42,44] on the other can be resolved. The scrambling of the labeling pattern when (22) is incorporated into tropine (1) can still only be explained by supposing that both enantiomers of the administered (22) are converted into (1). Nonstereospecific labeling of (1) from the amino acids could still be rationalized either by invoking putrescine as an intermediate on the pathway or by postulating racemization and subsequent incorporation of both enantiomers of an intermediate.

This analysis suggests that experiments aimed at proving the acetonedicar-boxylate route to (1) in Datura will have to assess the stereospecificity of incorporation of (4) and (5) into (1) and the pattern of incorporation of [1,2-13C2]ace-tate into (1) simultaneously. A further challenge will be to design and carry out successfully an experiment which will allow one to decide whether the C3 unit in the tropanes is elaborated via Path D or the acetonedicarboxylate pathway even if racemic intermediates are involved. It will be crucial to avoid the use of precursors that are as highly activated as the ^-keto acids fed by Leete [47,48] and Robins [44] and coworkers. This is because of the serious risk that subsequent incorporation of such putative precursors will be overinterpreted as being of mechanistic significance when in actual fact the result of an artefact is being observed. Thus for instance, feeding labeled acetonedicarboxylic acid to intact plants of the genus Datura or root cultures derived therefrom will not be the appropriate way to prove this notion. Under such circumstances we may expect that a Robinson type condensation may occur if a sufficient amount of (10) and a monoamine oxidase activity are present within the plant. Such a partially non-biological sequence would be expected to yield (18) which is also an intermediate in the biological sequence to (1).A comparison of the specific incorporation of (22) into cuskhygrin (32) (9% above natural abundance) vs (2) (2% above natural abundance) during a recent feeding experiment in Datura stramonium [44] may serve as a cautionary tale.

The situation in our work on lycopodine (27) was different and a feeding experiment with acetonedicarboxylic acid could be more easily justified. Firstly, more meaningful tracer evidence from incorporation of acetate into (27) was available which was suggestive of a role of acetonedicarboxylic acid in the pathway to lycopodine. Secondly, a "biomimetic" formation of lycopodine from acetonedicarboxylic acid and imine (14) in the presence of ubiquitous enzymes such as a monoamine oxidase cannot be readily envisaged given the greater structural complexity of (27) when compared to (18).For this reason we are confident that the incorporation of acetonedicarboxylic acid into (27) reflects the biological process in Lycopodium species.

0 0

Post a comment