Total solid-phase synthesis of the mixirins; antitumour cyclic lipopeptides

Mixirins A, B and C were recently isolated from a marine Bacillus sp. bacterium and have shown cytotoxic activity against human colon tumour cells. The mixirins were identified as a mixture of cyclic lipopeptides belonging to the iturin family. The iturins contain seven alpha-amino acid residues (five proteinogenic and two D-amino acids) and one unusual lipophilic beta-amino acid residue. Until now, the mixirins have yet to be synthesised and have their pharmacotherapeutic potential fully explored. This report describes the development and optimisation of novel methodology to enantioselectively synthesise lipophilic beta-amino acids, and demonstrates this through the syntheses of the beta-amino acid constituents of the mixirins. This methodology can also be used as a general procedure for homologation of alpha-amino acids to the corresponding beta-amino acids. The synthesised beta-amino acids were then utilised in the first reported total syntheses of mixirins A, B and C using established SPPS techniques. The absolute configuration of the beta-amino acid component of the mixirins (which had not been reported) was suggested to be of the (R)-configuration from macrolactamisation studies. Eight novel compounds with potential anticancer activity have been synthesised: mixirins A, B and C, and the beta-amino acid epimer of each, along with two mixirin analogues containing a beta-amino acid residue of reduced lipophilicity. Early biological testing has demonstrated promising results, confirming their pharmacotherapeutic potential. A study was carried out to determine whether the biosynthetic site of lipopeptide cyclisation is the most efficient for use during SPPS. Iturin A was chosen as it is a cyclic lipopeptide possessing significant biological activity, its biosynthesis via NRPSs is well-studied, and its total solid-phase synthesis has not been reported. Out of six trials (each with a different ring-closure site), only four proved suitable for the routine and high yielding synthesis of iturin A. The remaining two proved less convenient due to difficulties in the initial residue side-chain immobilisation. Ring closure between D-Tyr-3 and Asn-2 proved most efficient. Further work should be undertaken to pursue this interesting hypothesis and to investigate whether SPPS can be optimised from biosynthetic knowledge.