Patulin hemiacetal

Patulin (348) is a quite small natural product, but it nonetheless has a diverse complement of potent chemical functionality cyclic hemiacetal (the natural product is racemic) and allylic acetal moieties are present, and the molecule potentially can react with the nucleophihc residues of biomolecules as either a 1,4- or 1,6-Michael acceptor (see Fig. 7.1). Neopatulin (349) (296), ascadiol (350), and penicillic acid (351) are structurally related ylidenebutenolides, as are also metabolites of tetronic acid (352). 

The initial synthesis of Woodward and Singh was reported for patulin (348) in 1950 (297). However, the yield was low and the synthesis impractical. Two almost identical syntheses were reported independently by the groups of Stapleton in 1988 298, 299) and Riguera in 1989 300). Both commenced from L-arabinose, a readily available starting material from the chiral pool notwithstanding this, deprotection of a late-stage acetal intermediate to deliver the natural product was followed, perhaps unsurprisingly, by rapid racemization at the hemiacetal center. For details of these earlier syntheses, the reader is directed to our recent review (70) and the primary literature. 

Patulin (Fig. 1) was discovered by Birkinshaw etal. (1943), and its chemical structure elegantly proved and synthesized by Woodward and Singh (1949, 1950). The oxime was independently prepared by another route (Serratosa, 1961). A recent crystallographic study by Hubbard et al (1977) confirmed the structure of patulin and gave the bond distances and angles of the patulin molecule. Patulin isolated from natural sources has one asymmetric center on carbon 1. Patulin is, however, racemic. One explanation could be that the center at C(l) may racemize easily during work-up. Another possibility could be that the hemiacetal formation is not enzyme controlled. 

Fig. 40. Postulated mechanism of the conversion of the seven-membered ring hemiacetal to patulin.

---