Towards the total synthesis of amphidinolides C and F

The amphidinolides are marine macrolides extracted from dinoflagellates of the genus Amphidinium. To date, 37 amphidinolides have been isolated and identified, most of them possessing cytotoxicity against human cancer cell lines. Among these, amphidinolides C, F, C2 and C3 represent synthetic targets of interest owing to their scarcity, structural complexity and promising biological activities. This thesis describes the work realised towards the total synthesis of amphidinolides C and F, with a focus on the different strategies investigated and the key fragments synthesised. In the first approach, the C18−C29 fragment of amphidinolide F was prepared using an intramolecular etherification of an epoxide under acidic catalysis to produce the 2,5-trans-disubstituted tetrahydrofuran ring featured in the natural product. Unfortunately, dithiane alkylation with the C1−C17 iodide counterpart generated the desired coupling product in low yield. A second approach proposing to build the C17−C18 bond by a silicon-tethered RCM proved unsuccessful, because the requisite diene could not be obtained. It was then envisioned to form the C18−C19 bond by displacement of a triflate with an alkyne and install the ketone at C18 by a protoborylation/oxidation sequence. To this end, the C19−C29 triflate precursor was synthesised. Displeasingly, the C1−C18 alkyne counterpart (work by Dr Filippo Romiti) could not be prepared and coupling of the two fragments was not attempted. In the latest approach, the C10−C29 fragment of amphidinolide F was obtained employing a boron-mediated aldol condensation and a dithiane alkylation to form the C13−C14 and C18−C19 bonds. Several endgame strategies were examined including the successful Yamaguchi esterification of the C13-epi C10−C29 fragment and the C1−C9 acid. A challenging Stille crosscoupling was then effected to close the macrocycle but only yielded the desired macrolactone in trace amounts after global desilylation.