Two genes involved in the biosynthesis of the depsipeptide antibiotics pristinamycins I (PI) produced by Streptomyces pristinaespiralis were cloned and sequenced. The 1.7-kb snbA gene encodes a 3-hydroxypicolinic acid:AMP ligase, and the 7.7-kb snbC gene encodes PI synthetase 2, responsible for incorporating L-threonine and L-aminobutyric acid in the PI macrocycle. snbA and snbC, which encode the two first structural enzymes of PI synthesis, are not contiguous. Both genes are located in PI-specific transcriptional units, as disruption of one gene or the other led to PI-deficient strains producing normal levels of the polyunsaturated macrolactone antibiotic pristinamycin II, also produced by S. pristinaespiralis. Analysis of the deduced amino acid sequences showed that the SnbA protein is a member of the adenylate-forming enzyme superfamily and that the SnbC protein contains two amino acid-incorporating modules and a C-terminal epimerization domain. A model for the initiation of PI synthesis analogous to the established model of initiation of fatty acid synthesis is proposed.Pristinamycins I (PI), cyclohexadepsipeptide antibiotics produced by Streptomyces pristinaespiralis, are members of the streptogramin B group. They are coproduced with the polyunsaturated macrolactone antibiotics pristinamycins II (PII), members of the streptogramin A group (6, 52). Like many other small metabolites with a peptide structure, PI are synthesized nonribosomally by large multifunctional enzymes (49). More than 25 years ago, Lipmann emphasized the analogy between nonribosomal peptide synthesis and fatty acid synthesis (25). In his model, large multifunctional enzymes called peptide synthetases (PPSs) catalyze elongation of a peptide covalently linked to a phosphopantetheinyl arm by a thioester bond (for a review, see reference 23). The original version of the thiotemplate multienzymatic mechanism was revised recently, and a multiple-carrier model was proposed (39,43,46,53). According to this model, each amino acid is activated as an aminoacyl adenylate and linked to the enzyme as a thioester with a phosphopantetheinyl group. Elongation then occurs by transfer of the activated carboxyl to the amino group of the next amino acid, thus effecting N-to-C stepwise condensation. Primary structure analysis of several PPS genes resulted in the identification of approximately 1,000-aminoacid (aa)-long modules selectively catalyzing activation and condensation of one amino or hydroxy acid (30, 51). The modules are organized such that they are colinear with the sequence of the oligopeptide. Each module can be subdivided in domains. The activation domain (500 aa) belongs to the large family of adenylate-forming enzymes that includes firefly luciferase and acyl coenzyme A (acyl-CoA) synthetases and contains nine core sequences (named boxes A to I by Pfeifer et al.

SummaryFour pap genes (papA, papB, papC, papM ) were found by sequencing near to snbA, a Streptomyces pristinaespiralis gene which was previously shown to encode one of the pristinamycin I (PI) synthetases. Analysis of the homologies observed from the deduced amino acid sequences suggested that these four genes could be involved in the biosynthesis of the PI precursor 4-dimethylamino-L-phenylalanine (DMPAPA). This was first verified when disruption of papA in S. pristinaespiralis led to a PI ¹ phenotype, which was reversed by the addition of DMPAPA into the culture medium. Further confirmation was obtained when papM was overexpressed in Escherichia coli and the corresponding protein purified to homogeneity. It catalysed the two successive N-methylation steps of 4-amino-L-phenylalanine leading to DMPAPA via 4-methylamino-L-phenylalanine. These results allowed us to assign a function to each of the four pap genes and to propose a biosynthetic pathway for DMPAPA.

No abstract

N. BAMAS‐JACQUES, S. LORENZON, P. LACROIX, C. DE SWETSCHIN and J. CROUZET.1999.Streptomyces pristinaespiralis synthesizes pristinamycin, a member of the streptogramin antibiotic family which consists of a mixture of two types of chemically unrelated compounds named pristinamycins I and pristinamycins II. In order to estimate the size of the Strep. pristinaespiralis chromosome and to elucidate the organization of the pristinamycin biosynthetic and resistance genes already identified, it was decided to use the pulsed‐field gel electrophoresis technique. Results indicate that the Strep. pristinaespiralis chromosome is linear and about 7580 kb, as previously shown for several other Streptomyces species. By hybridization, it could be shown that the biosynthetic and resistance genes for pristinamycins I and pristinamycins II, except for the multidrug resistance gene ptr, are interspersed and seem to be organized as a single large cluster, covering less than 200 kb corresponding to 2·6% of the total size of the chromosome. The consequences and significance of such a genetic organization are discussed.