Towards the total synthesis of the cytotoxic neo-clerodane salvileucalin B

Abstract

The work described in this Thesis was directed towards establishing a total synthesis of the cyclopropane-containing natural product salvileucalin B (1.3). The initial focus was the formation of its caged core substructure as embodied in lactone 1.13. This incorporates an alpha-methylene-gamma-butyrolactone-type subunit thought to be contributing to the biological properties of the natural product. The protecting group free synthesis of compound 1.13 was achieved using an intramolecular Buchner reaction as the key step. Elaboration of the [4.3.1]propellane 2.37 formed as a result of this process engaged in unusual and reversible rearrangement reactions leading to novel fenestranes. The enantioselective desymmetrisation of meso-compound 1.13 was achieved through Kornblum-DeLaMare rearrangement of the derived endo-peroxide, thus providing the means by which an enantioselective total synthesis of target 1.13 could eventually be realised. Chapter 1 provides a brief introduction to the most important compounds isolated from plants of the Salvia genus, including the title compound 1.3. This is followed by a discussion of previous efforts directed towards its synthesis. Chapter 2 details the synthetic work carried out by the author in seeking to prepare lactone 1.13. The original approach, involving a dibromocarbene addition reaction to establish the [4.3.1]propellane framework, was eventually replaced by an intramolecular Buchner reaction that exploited the rhodium-catalysed decomposition of diazoketone 2.34. Both approaches used 2-indanone 2.7 as the starting material. Once the target core lactone 1.13 was successfully synthesised it was subjected to hydrogenation reactions employing three different types of catalysts. In this way the partially and fully hydrogenated lactone analogues 2.77 - 2.82 were produced with three of these still containing the alpha-methylene-gamma-butyrolactone-type subunit. These compounds and the core lactone 1.13 itself were each subjected to biological testing which revealed, surprisingly, that they were all essentially inactive. Chapter 3 details investigations into a rearrangement that was discovered during the course of carrying out the DIBAL-H-mediated reduction of the nitrile group within compound 2.61 and which led, unexpectedly and presumably via a [3,5]-sigmatropic rearrangement of aldehyde 2.62, to the formation of oxa-[5.6.5.6]fenestratetraene 3.51. Computational calculations support the notion that the conversion of 2.62 into 3.51 doesn't involve a [3,3]-sigmatropic rearrangement followed by a 1,3-shift. Subjection of fenestrane 3.51 to a second DIBAL-H-mediated reduction yielded the symmetrical alcohol 2.63, thus confirming the reversible nature of the proposed [3,5]-sigmatropic rearrangement process. Chapter 4 describes the investigation into the use of the Buchner reaction as a means for making various propellanes and fenestranes related to compounds 1.13 and 3.51, respectively. However, such efforts were thwarted by a range of C-H insertion reactions that occurred instead of the desired intramolecular Buchner process. The work reported in Chapter 5 was concerned with the late-stage desymmetrisation of meso-core lactone 1.13. Several desymmetrisation protocols were investigated with the enantioselective Kornblum-DeLaMare rearrangement of the derived endo-peroxide 5.8 being the most effective and forming gamma-hydroxyenone 5.10. A possible synthetic route to salvileucalin B (1.3) based on such work is presented.