(2S)-trans-4-Hydroxyproline has been studied as a chiral template for the preparation of kainoids. Conversion to protected 4-ketoproline derivatives has allowed functionalisation at C-3 to be investigated. Although the tendency to epimerisation has caused problems, we have successfully introduced a variety of alkyl side chains with cis stereochemistry at this position. Functionalisation of the 3-alkyl-4-ketoprolines at C-4 has been investigated and a method has been found to prepare compounds which are functionalised stereospecifically at both C-3 and C-4. These have potential for study as kainate analogues. During our studies on the synthesis of kainoids, an unusually stereospecific deuteriation has been discovered. We have exploited this in a synthesis of (2S,3S)-[3- 2Ht]-proline, (2S,3R)-[3-2Ht]-proline and (2S)-[3,3-2H21-proline.

Attempts were made to synthesise a conformationally restricted analogue of kainic acid wherein the double bond was confined in a ring-system. The stratagem involved an intramolecular Diels-Alder reaction but could not be tested as the precursors to the cycloaddition reaction could not be prepared. Attempts were made to develop a general route to kainic acid and analogues by employing a 1,3-dipolar cycloaddition reaction between aziridines and olefins. Triazolines were used as a precursor to aziridines because of the ease of formation from alkyl azides and olefins. The required dipolar cycloaddition was found to occur but produced various side-products from the triazoline thermolysis. The subsequent Grignard reaction on the cycloaddition product gave problems as the compound epimerised under basic conditions and did not undergo reaction with methyl Grignard or methyl lithium. An attempt to prepare kainic acid and analogues by an intramolecular 1,3-dipolar cycloaddition or a 1,3-sigmatropic shift reaction failed when the basic precursors for the reaction could not be prepared.