Fermentation-derived compounds as a source of new products

New products may be sought by modifying older fermentation products, exemplified by the semisynthetic antibiotic tilmicosin and other derivatives of tylosin. Another approach is to screen fermentation broths, exemplified by the discovery of the structurally-unique insecticidal macrolide spinosad. The polyketide class of fermentation products encompasses a wide diversity of chemical structures and biological activities. Important products include macrolide antibiotics (erythromycin, josamycin, tylosin), antifungal polyenes (amphotericin B), parasiticides (avermectin, milbemycin), and immunosuppressive agents (FK-506, rapamycin). The past history of success achieved by this class suggests it will remain an important source of new products. The discovery of new products may be divided into three approaches: 1) structural modifications of previously known compounds that improve or extend features of the parent molecules, 2) screening culture broths to identify new biologically-active compounds, and 3) manipulation of the biosynthetic pathways by which secondary metabolites are formed in order to produce new structures. Semisynthetic derivatives of tylosin: Mutants strains of Streptomyces fradiae initially provided several new macrolides as starting materials for chemical modification [ 11, illustrating the benefits to natural product chemists from collaborations with microbiologists. Analysis of structure-activity relationships among these biosynthetic intermediates, shunt metabolites, and their initial semisynthetic derivatives had presented two contrasting directions for further modification: 1) derivatives of 5-0-mycaminosyltylonolide (OMT), especially at C-23, had potent in v i m antibiotic activity and in vivo efficacy after parented administration in animal models of infection, but relatively poor oral bioavailability, whereas 2) modifications of the aldehyde in tylosin and desmycosin substantially improved the ratio of ordparenteral efficacy and bioavailability [2]. This contrast in research direction emphasized the critical importance of basing SAR studies on in vivo activity (especially oral activity) rather than using only in vitro activities to guide S A R studies. Pursuit of the latter direction toward some well-defined objectives necessary for a new veterinary antibiotic ultimately led to tilmicosin, initially developed to treat bovine respiratory disease by a single injection due to its long and persistent in vivo halflife [Micotil(R)], and later developed as a feed additive [Pulmotil(R)] to control pneumonia in pigs [2]. Several new modifications of desmycosin have been subsequently explored, including: 1) modification and replacement of mycinose; 2) modification of the dienone moiety; 3) modification of mycaminose; and 4) modification of the C-3 hydroxyl group. Many of these modifications were combined with modifications of the aldehyde in order to incorporate anticipated pharmacokinetic benefits of modifying the latter, as observed in tilmicosin and other aldehyde-modified macrolides. One goal of this effort was to find ways to inhibit resistant bacteria. From the first approach, modifications of mycinose or its replacement by deoxysaccharides or by numerous non-saccharide substituents in 20-modified derivatives of desmycosin had not revealed improvements in antibacterial potency, spectrum, or efficacy in animal models of infection. An intriguing result was in v i m antifungal activity exhibited by certain derivatives, but this unexpected finding was not further developed due to lack of in vivo efficacy against fungal infections in mice and some mammalian toxicity [3]. 12,13-Epoxydesmycosin was synthesized by protection of the aldehyde (EtOH, pTSA), simultaneous epoxidation and N-oxide formation (MCPBA, CHC13, O"), selective reduction of the