The Biosynthesis of Sulfomycin Elucidated by Isotopic Labeling Studies

• Published in: Journal of the American Chemical Society 1996-11, Vol.118 (46), p.11363-11368
• Main Authors: Fate, Gwendolyn D, Benner, Charles P, Grode, Stephen H, Gilbertson, Terry J
• Format: Article
• Language: English
• ISSN: 0002-7863; 1520-5126
• DOI: 10.1021/ja961864g

U-102408 has been isolated from S. arginensis cultures and fully characterized by NMR and MS. These studies indicate that U-102408 is identical with sulfomycin, and this data confirm the structure of sulfomycin which was assigned by chemical degradation and limited NMR studies. The biosynthetic origins of all U-102408 structural features have been elucidated through isotopic labeling experiments with primarily 13C but also using 14C, deuterium or tritium. Of particular interest, the 4-hydroxy-2-amino-2-pentenoic acid moiety was determined to originate from threonine and a one carbon unit derived from the two position of glycine or the S-methyl of methionine. The biosynthetic pathway for U-102408 differs from that for nosiheptide, thiostrepton, and berninamycin in that cysteine and serine do not cross label sites in U-102408, while this has been observed for the other thiopeptides. In U-102408 cysteine and serine are each incorporated into unique sites within the molecule. This indicates that the interconversion of cysteine and serine is not an active pathway for S. arginensis. In fact for fermentations in defined medium the addition of cysteine was found to be necessary to achieve higher antibiotic titer. Unlike the other thiopeptides that have been studied, threonine was incorporated into sites labeled by serine though serine did not incorporate into sites labeled by threonine within U-102408. No cross labeling of threonine and serine has been observed in studies of nosiheptide, thiostrepton, or berninamycin. In the current study glycine is investigated as a precursor for U-102408. The interconversion of glycine to serine is a very facile pathway in S. arginensis; glycine labels all sites labeled by serine. This data suggests glycine as a possible precursor for nosiheptide, thiostrepton, and berninamycin which has not been investigated for these other thiopeptides. Incorporation of label from 3-2H-serine is facile for thiostrepton and nosiheptide and was used to probe the mechanism of formation of the pyridine ring. However, no incorporation from 3-2H-serine or 3-3H-serine was observable for U-102408. While the lack of incorporation may be due to washout of the label from the precursor, it also may be due to the interconversion of serine and glycine causing loss of the label from the serine pool.