Erythronolide A glycosidation to erythromycin A by a blocked mutant of Streptomyces erythraeus.

The commonly accepted biosynthetic pathway1) leading to erythromycin A (I) requires that substrate for sugars (L-mycarose and Ddesosamine) attachment should be erythronolide B (IV). Once both sugars are linked to the aglycone, giving rise to erythromycin D, C-12 hydroxylation would occur resulting in erythromycin C production. Erythromycin C is further converted to erythromycin A by O-methylation of the neutral sugar (at C-3"). According to this scheme, erythronolide A (III) is not considered to be a biosynthetic intermediate. This view is supported by the fact that neither crythronolide A nor its neutral monoglycoside were ever detected and isolated in the fermentation broths of normal strains or blocked mutants of Streptomyces erythraeus. On the contrary the corresponding 12-deoxy compounds were found2,3). We investigated whether permeation barriers or strict substrate specificity could prevent conversion of erythronolide A to erythromycin A in Streptomyces erythraeus. Bioconversion studies using blocked mutants are often useful for this purpose, since it is known that other modified lactones, related to erythronolide B, often accept L-mycarose, giving rise to novel monoglycosides which, with a few exceptions1,4), usually do not serve as acceptors for D-desosamine2 . During our investigations, we attempted to feed erythronolide A, obtained by glycoside cleavage reaction on erythromycin A8), to a suitable early blocked mutant of a high erythromycin producing industrial strain, incapable of synthesizing the antibiotic de novo, but able to convert biosynthetic precursors to erythromycin. The strain Pierrel LMC 1648 (registered as ATCC 31772) was derived from Streptomyces erythraeus Pierrel LMC 1056, an erythromycin producer, by UV irradiation. It was selected as a blocked mutant by the agar-disc method'), and its ability to convert erythronolide B to erythromycin A (I) and erythromycin B (II) in variable amounts was assessed by preliminary shake flask trials. The culture for microbial conversion of crythronolide A was prepared as follows: strain ATCC 31772 was inoculated into 100 ml of seed medium consisting of 3.0% sucrose, 0.8% corn steep solids, 0.9 % soybean oil, 0.2 % ammonium sulfate, 0.7 % calcium carbonate, in a 500-m1 Erlenmeyer flask. Following incubation for 2 days at 33°C under shaking at 220 rpm, 1.5 ml of the seed culture were inoculated into 30 ml of a medium containing 3.0% corn dextrins, 4.0 raw corn starch, 3.0% soybean meal, 2.0 soybean oil, 0.2% ammonium sulfate, 0.6 calcium carbonate, in a 250-ml Erlenmeyer flask at 33°C for 120 hours. To the culture was added 500 /cg/ml of crystalline erythronolide A 24 hours after the start of fermentation. Erythronolide A was detectable in the culture filtrate up to 16 hours after the addition. It disappeared with rapid conversion to erythromycin A alone, as proven by the HPLC procedure of Tsu.I1 et