Synthetic studies towards the spiroimine unit of the spirolides

Abstract

The spirolides (A-D) were isolated from the digestive glands of mussels, scallops and toxic plankton from the east coast of Nova Scotia in Canada. These toxins are metabolites of the dinoflagellate Alexandrium ostenfeldii and A. peruvianum. To date fourteen members of this family have been identified from around the world. The spirolides are considered fast acting neurotoxins due to their effect in the mouse bioassay and activate L-type calcium channels. They also represents the most potent non-peptidic nicotinic acetylcholine antagonists, rendering them attractive candidates for the treatment of cardiovascular disorders and candidates for the development as antagonists of specific receptor sub-types. The total synthesis of the spirolides has not yet been reported. Their structure can be divided in two parts; an upper novel 6,7-spiroimine unit and a lower 5,5,6- bis-spiroacetal moiety which was previously synthesised by the Ishira and Brimble groups. However, the spiroimine unit has not yet been synthesised, representing a synthetic challenge of great interest due to the fact that this part has been established as the putative pharmacophore of this family of marine biotoxins. This study provides two approaches to the bicyclic spiroimine moiety of the spirolides A and B. The first proposed retrosynthesis accesses the C29 quaternary stereocentre using a highly diastereoselective Birch reductive alkylation. Further studies on the diastereoselective hydrogenation of the resulting double bonds, cleavage of the auxiliary and cyclisation studies afforded a range of enantiopure bicyclic spirolactams. During these studies, a synthetic approach to the challenging A ring of the spirolides A and B has been established via diastereoselective Myers alkylation of an amide derived from (S,S)-pseudoephedrine. Unfortunately, the lactam functionality could not be converted to an imine. Subsequently, an improved retrosynthetic pathway, making use of a Diels-Alder reaction to simultaneously establish the C7 and C29 stereocentres of the spirolides was extensively studied. Further transformation of the Diels-Alder adduct afforded the advanced azido-aldehyde or azidoketone in good overall yield. Studies of the cyclisation using various phosphines with the two different substrates afforded the surprising crystallographic structure of a stable 14-membered diimine ring and also the successful synthesis of the 6,7-spiroketimine unit of the spirolide. This new strategy has lead to a better understanding of the stability of the imine unit and allowed the development of a viable pathway for the synthesis of the 7,6-spiroimine unit of the spirolides.