Synthesis of the Tricyclic Core of Vinigrol via an Intramolecular Diels-Alder Reaction

Link: http://pubs.acs.org/cgi-bin/abstract.cgi/orlef7/asap/abs/ol0702977.html

From Prof. Louis Barriault's group at the University of Ottawa, Ottawa, Canada

A recent article in Org Lett ASAP detailing a partial total synthesis of vinigrol (1), particularly the tricyclic core structure of the natural product. The natural product shows antihypertensive and platelet aggregation-inhibiting properties.

The authros planned to use intramolecualr Diels-Alder reaction to construct this core. The planned cycloaddition looked quite funky as the orientation of the dienophile seemed quite strained in order to have a good interaction with the diene, not to mention that the orientation of diene had to cooperate. Nonetheless, the strategy seemed ambitious. In their retrosyn, they were led back to aldehyde 10 as the simpler building block.

Another aspect of the cycloaddition that required attention was the different approach of the dienophile to give a regioisomeric product 13 as shown in Scheme 3. Although the transition state leading to cycloadduct 13 seemed unlikely, this partial synthesis would also serve as a confirmation of this hypothesis.

In the forward direction, the synthesis started with aldehyde 10 as shown below.

Takai elefination of aldehyde 10 gave vinyl iodide 15. Buchwald's coupling protocol of 15 with 16 gave the ether 17. The care with temperature had to be taken as slightly higher temperature than 90 C would increase the amount of aldehyde 18 with epimerization at the alpha carbon. After ether 17 was obtained, it was subjected to i-Bu3Al as the Lewis acid to promote stereocontrol [3,3] sigmatropic rearrangement followed by immediate reduction to alcohol 19. Silylation then afforded 20.

Alkene 20 was subjected to conditions in scheme below. A more direct synthesis of the nitrile 22 would be to use of Grubbs' catalyst to perform cross metathesis with acrylonitrile followed by hydrogenation. But this did not work; only SM was returned. So it was resorted to KCN displacement of OTs group obtained from hydroboration-oxidation and then tosylation of alkene 20.

Then alkyne 24 was synthesis from the corresponding aldehyde of 23 using modified Ohira's protocol ((a) Roth, G. J.; Muller, S.; Bestmann, H. J. Synthesis 2004, 59. (b) Ohira, S. Synth. Commun. 1989, 19, 561.) as the aldehyde is sensitive to epimerization. The Wittig olefination of aldehyde 25 to alkene 26 was performed using Conia conditions (Conia, J.-M.; Limasset, J.-C. Bull. Soc. Chim. Fr. 1967, 114, 1936.)

An always-cool enyne ring-closing metathesis using Grubbs' second generation catalyst then afforded diene 27. Attempts to directly convert nitrile to corresponding enone 11 using Grignard reagents and various additives failed. Therefore, it was resorted to step-wise operations as shown.

BF3-OEt2-catalyzed Diels-Alder reaction then afforded the desired cycloadduct 12 in almost quantitative yield (as the only regioisomer). This result was not surprising as DFT calculations using Khon-Sham DFT at the B3LYP19 level of theory with a 6-31G** basis set also confirmed that the transition state leading to regioisomer 13 was 10.7 kcal/mol higher than the one leading to 12 (Scheme 7).

An ok partial synthesis overall with the exception of some noteworthy steps. The synthesis was also a little too linear.