Stereocontrolled Total Synthesis of (-)-Kainic Acid

Link: http://pubs.acs.org/cgi-bin/abstract.cgi/orlef7/asap/abs/ol0631197.html

From Prof. Tohru Fukuyama's group at University of Tokyo, Tokyo, Japan

A new stereoselective synthesis of (-)-kainic acid appeared in Org Lett ASAP. This time it came from Prof. Tohru Fukuyama in Japan. Kainic acid is a parent member of the kainoid family. Kainoids display potent anthelmintic properties and neurotransmitting activities in the mammalian central nervous system, and kainic acid in particular has been widely used as a tool in neuropharmacology for stimulation of nerve cells and the mimicry of disease states such as epilepsy, Alzheimer’s disease, and Huntington’s chorea.

The general plan for the forward direction is as followed.

In the retro, kainic acid (1) was traced back to oxazolone 6, through a series of transformations including the key 1,4-addition of enolate to enone lactone, which was to be built by RCM.

Therefore the synthesis commenced with construction of RCM precursor 13. The key reaction included crotyl aldol reaction of 8 to give 9 using Evan's aldol auxiliary, and Mitsunobu reaction to form 12.

But Mitsunobu reaction was not practical for larger scale synthesis to install the glycine fragment, a new route was devised. In this new route, the glycine fragment was installed using intermediate aminal 15 via reductive amination.

The low yielding step to make 16 (Scheme 3), was improved by starting with chiral auxiliary 17. Fragment 16 was further manipulated using the same sequence in Scheme 2 to give RCM precursor 13.

Hoveyda-Grubbs catalyst was chosen and conditions were screened for the best one. The reaction proceeded most optimally in dichloroethane with only 0.8 mol% catalyst loading (entry 10). The loading could be decreased to 0.5 mol% while maintaining relatively high RCM yield (entry 11).

The same enone 21 could be constructed using a different route in attempts to avoid the use of expensive RCM catalyst. The key steps in this sequence were constructions of enone fragment 23 and 24. Enone 23, obtained in 83:17 ratio of the desired Z-isomer, could cyclize directly to 21 in two steps. But for E-enone 24, the double bond was temporarily removed to facilitate cyclization before it was re-installed in the last step by oxidation of sulfide 25 to intermediate sulfone with ozone, followed by heating to eliminate sulfonic acid to give 21 in the total of three steps (Scheme 5).

The next key step was the 1,4-addition. In order to control the stereochemistry of the 2-position of pyrrolidine ring, several conditions were screened. It was found that LiHMDS in DMF worked best to give the best ratio of 26a/26b in excellent yield (Table 2 entries 5, 6, and 7). It should be noted that even though when R=tert-butyl group gave the best yield and ratio of 26a, the construction of t-Butyl derivative of ester 16 from 19 proceeded in very poor yield (see Scheme 4).

Fragment 26 was then subjected to methanolysis to give 27, followed by TPAP/NMO oxidation to ketone 28. The next key step was olefination of methyl ketone under non-basic conditions to prevent epimerization at C4 position (of the pyrrolidine ring) to give 29. Two further routine steps then afforded (-)-kainic acid.