No abstract

Zahlreiche funktionelle Gruppen enthaltende Ester (I) werden mit Hilfe von Titan(IV)‐alkoxiden in bekannter Weise umgeestert.

The ioacrodiotide antibiotic (+)-colletodiol (7, &home 1) was synthesiscd. The configurntion was thus proven 10 bc (6R.1 t R,12R.t4R), -Kcy intermediates (Schemes 9 and 10) are the two hydoxy acids 43 and 64, which were preparcd from dimethyl (S.S)-tartrate in 20% overall yield (12 s!eps) and horn IR)-3-hydroxybutanoate in 57% overall yield (6 steps), respectively. The two hydroxy acids were cyclised to give, aker deproicction. rhc title compound. -Our investigations led to the production of a largc number of chiral building blocks, many ofthcm in differcnt stercoisomcric forms. A) IntroductionTwo groups of medium-ring diolides (macrodiolides) have been isolated from different fungi. The unsymmetrical ones are shown in Scheme 1. A discussion of these diolides along with those of C2-symmetry (references to isolation, structure elucidation, and synthesis) was provided in a recent paper 'I. For the total synthesis of (+)-conglobatin see ref.6).We initiated a project directed toward the total synthesis of (+)-colletodiol(7) in order to demonstrate the utility of chiral building blocks derived from tartaric acid. A list of references to such building blocks is given in a review article') (see also ref. ' 1) .In the initial publication on the structure of (+)-colletodio19) misinterpretation of the available data led to the assignment of configuration at C-11 as S. Therefore we first attempted to synthesise what is now known as 5-epicolletodio13). This led to a number of 1,2-diols of u configuration") (see -13) ). When we compared an advanced intermediate of the synthesis with the corresponding compound derived from degradation of colletodiol, it was evident, that the published structure was wrong. An X-ray analysis showed that the 1,2-diol unit has 1 c~nfiguration'~). Thus the corresponding building blocks with 1 configuration were prepared.A description of the preparation and the full characterisation of all chiral building blocks derived from tartaric acid during this work is given in and in several publications (see ref. 7J-13)).Many of these epoxides and 1,2-diols are also available through Sharpless epoxidation ''I. For convenient laboratory-scale preparations of enantiomerically pure samples, the present "chiral pool" approach can compete very well with the enantioselective epoxidation. In the following sections, our total synthesis of (+)-colletodiol is described in detail. A short communication was published in ref. B) Synthesis of (+)-ColletodiolSeveral synthetic strategies were followed to synthesise colletodiol (7). The following connections were made to build up the target (Scheme 2): The C(lO)-C (13)

The macrocyclic pentolide 1, hexolide 2, and heptolide 3 constitute ca. 80% of the oligomers formed in ca. 50% yield from enantiomerically pure 3-hydroxybutanoic acid under Yamaguchi's macrolactonization conditions. The FAB mass spectra of the MH', MNaf, and MCsf are reported (Figs. 2, 3, 5, and 6 ) . No cyclic tetramer is detected. The 'H-NMR spectra of the cyclic oligomers, of the monomer, and of the polymer (PHB) are very similar (Fig. 4 ) . Directed synthesis of the open-chain dimer and tetramer of 3-hydroxybutanoic acid and attempted cyclization do not lead to the isolation of the cyclic tetramer.

No abstract