English

Cyclodepsipeptides of the enniation-, PF1022- and verticilide-family represent a diverse class of highly interesting natural products with respect to their manifold biological activities. However, until now no practicable solid-phase synthesis of these compounds has been accomplished, probably due to the problematic combination of N-methyl amino acids and hydroxycarboxylic acids. The first part of this work deals with the development of synthetic methods to get access to the building blocks needed for the solid-phase synthesis of the anthelmintic PF1022A, its commercial analogue emodepside and a library of PF1022- and bassianolide analogues. The building blocks except the (R)-(p-morpholinophenyl)-lactic acid (morphPhLac) were available via standard procedures. Subsequently two enantioselective methods were developed. One is using an enzymatic procedure consisting of a cofactor regenerating redox system working with the enzymes D-lactate dehydrogenase (D-LDH) and formate dehydrogenase (FDH). The enzymatic system reduces α-oxocarboxylic acid derivatives enantioselectively to the corresponding lactic acid derivatives. The cofactor (NADH) is regenerated by reducing NAD in the presence of FDH and a formate salt. The enzymes can be reused by using a membrane reactor. The other method is based on a metal catalyzed asymmetric hydrogenation to reduce enol esters enanti-oselectively. Both methods provide several derivatives in high yields and enantiomeric purities. In the second part the building blocks were used to develop different methods for a solid-phase synthesis of PF1022A, its commercial analogue emodepside and other PF1022- and bassianolide-analogues on Kaiser- and Wang-resin. One method, using didepsipeptide building blocks, could be developed on Kaiser- and Wang-resin. The Kaiser-resin has the advantage that a cyclizative cleavage of the linear depsipeptide to the final cyclooctadepsipeptide is possible. PF1022A and emodepside could be synthesized in excellent overall yields. For PF1022A it was also developed on Wang-resin. The Wang-resin was chosen because of its more stable linkage to the substrate. So it’s possible to choose the solid support, dependent on the properties of the building blocks needed for the cyclodepsipeptide derivative. In a second method single building blocks were used for a stepwise solid-phase synthesis of PF1022A on Wang-resin. This method allows structural variations in each step and it was possible to synthesize PF1022A, emodepside and three PF1022-analogues. Both methods provide the basis for the solid-phase synthesis of cyclodepsipeptide libraries with a high probability of anthelmintic, antibacterial or insecticidal activities.