| Summary: | This work continues previous studies carried out by the Jenkins group to produce taxanes from glucose and is split into two projects. The first investigates alkylation of cyclohexane-1,2-dicarboxylic acid mono(-)menthyl ester by deprotonation with LDA and treatment with a range of alkyl bromides and an iodide. The resulting crude products were converted to the corresponding methyl esters using diazomethane. (Fig. 8719A). In all cases both diastereoisomers were seen, but the stereoselectivity ranged from 7:1 to 32:1 depending upon the alkyl halide used. When the acid ester was treated with LDA followed by BnBr the major isomer isolated, after conversion to the methyl ester, was confirmed as the 1(R), 2(S) diastereoisomer by X-ray crystallography. The results of these stereoselectivity studies were compared to the ratios of alkylation products from treatment of cyclopentane-1, 2-dicarboxylic acid mono (+) menthyl ester with various alkyl halides as described in the literature. The aim of this project was to provide a C-ring with a methyl group at C-1 of the correct stereostructure for incorporation into a taxane skeleton - this could be achieved via a radical pathway based on a C -> ABC strategy. The second project investigated the addition of 3-bromo-2,4-dimethylpenta-1,3-diene to aldehydes. The diene was first treated with nBuLi to give a lithium-halogen exchange. Reaction with cyclohexane carboxaldehyde gave a mixture of the diene alcohol and allene alcohol in the ratio of 2.3:1. Addition of the lithiated diene to a suspension of CeCl3 before introduction to the aldehyde improved the ratio to >100:1. Other lanthanide (III) chlorides were also studied. (Fig. 8719B). Reaction of the lithiated diene to 2-benzoyloxy-2-(cyclohexyl)-ethanal in the presence of various lanthanide chlorides gave a mixture of both enantiomers of two diastereoisomers. The ratio of these diastereoisomers was dependent upon the nature of the lanthanide. CeCl3 was found to give the highest ratio. Using the knowledge gained from these model studies, it was attempted to add the lithiated diene in the presence of CeCl3 to the aldehyde group of a highly functionalised C-ring synthesised from glucose. Successful addition was not achieved, the reasons considered for this failure were local and distant steric hindrance. Alternative protecting groups to the bulky t butyldiphenylsilyl groups were sought and regression several steps back along the route allowed the introduction of small methyl protecting groups. The protected diol was advanced to a functionalised C-ring. Treatment with the lithiated diene in a suspension of CeCl3 allowed successful addition of the diene to the aldehyde. This product provided 4 of the 6 A-ring carbons and completed the top half of the B-ring. (Fig. 8719C).
|
|---|
