Comparative characterization of the lactimidomycin and iso-migrastatin biosynthetic machineries revealing unusual features for acyltransferase-less type I polyketide synthases and providing an opportunity to engineer new analogues.

Lactimidomycin (LTM, 1) and iso-migrastatin (iso-MGS, 2) belong to the glutarimide-containing polyketide family of natural products. We previously cloned and characterized the mgs biosynthetic gene cluster from Streptomyces platensis NRRL 18993. The iso-MGS biosynthetic machinery featured an acyltransferase (AT)-less type I polyketide synthase (PKS) and three tailoring enzymes (MgsIJK). We now report cloning and characterization of the ltm biosynthetic gene cluster from Streptomyces amphibiosporus ATCC 53964, which consists of nine genes that encode an AT-less type I PKS (LtmBCDEFGHL) and one tailoring enzyme (LtmK). Inactivation of ltmE or ltmH afforded the mutant strain SB15001 or SB15002, respectively, that abolished the production of 1, as well as the three cometabolites 8,9-dihydro-LTM (14), 8,9-dihydro-8S-hydroxy-LTM (15), and 8,9-dihydro-9R-hydroxy-LTM (13). Inactivation of ltmK yielded the mutant strain SB15003 that abolished the production of 1, 13, and 15 but led to the accumulation of 14. Complementation of the ΔltmK mutation in SB15003 by expressing ltmK in trans restored the production of 1, as well as that of 13 and 15. These results support the model for 1 biosynthesis, featuring an AT-less type I PKS that synthesizes 14 as the nascent polyketide intermediate and a cytochrome P450 desaturase that converts 14 to 1, with 13 and 15 as minor cometabolites. Comparative analysis of the LTM and iso-MGS AT-less type I PKSs revealed several unusual features that deviate from those of the collinear type I PKS model. Exploitation of the tailoring enzymes for 1 and 2 biosynthesis afforded two analogues, 8,9-dihydro-8R-hydroxy-LTM (16) and 8,9-dihydro-8R-methoxy-LTM (17), that provided new insights into the structure-activity relationship of 1 and 2. While 12-membered macrolides, featuring a combination of a hydroxyl group at C-17 and a double bond at C-8 and C-9 as found in 1, exhibit the most potent activity, analogues with a single hydroxyl or methoxy group at C-8 or C-9 retain most of the activity whereas analogues with double substitutions at C-8 and C-9 lose significant activity.

iso-Migrastatin, migrastatin, and dorrigocin production in Streptomyces platensis NRRL 18993 is governed by a single biosynthetic machinery featuring an acyltransferase-less type I polyketide synthase.

iso-Migrastatin and related glutarimide-containing polyketides are potent inhibitors of tumor cell migration and their implied potential as antimetastatic agents for human cancers has garnered significant attention. Genome scanning of Streptomyces platensis NRRL 18993 unveiled two candidate gene clusters (088D and mgs); each encodes acyltransferase-less type I polyketide synthases commensurate with iso-migrastatin biosynthesis. Both clusters were inactivated by lambda-RED-mediated PCR-targeting mutagenesis in S. platensis; iso-migrastatin production was completely abolished in the DeltamgsF mutant SB11012 strain, whereas inactivation of 088D-orf7 yielded the SB11006 strain that exhibited no discernible change in iso-migrastatin biosynthesis. These data indicate that iso-migrastatin production is governed by the mgs cluster. Systematic gene inactivation allowed determination of the precise boundaries of the mgs cluster and the essentiality of the genes within the mgs cluster in iso-migrastatin production. The mgs cluster consists of 11 open reading frames that encode three acyltransferase-less type I polyketide synthases (MgsEFG), one discrete acyltransferase (MgsH), a type II thioesterase (MgsB), three post-PKS tailoring enzymes (MgsIJK), two glutarimide biosynthesis enzymes (MgsCD), and one regulatory protein (MgsA). A model for iso-migrastatin biosynthesis is proposed based on functional assignments derived from bioinformatics and is further supported by the results of in vivo gene inactivation experiments.

TLN-05220, TLN-05223, new Echinosporamicin-type antibiotics, and proposed revision of the structure of bravomicins(*).

The deposited strain of the hazimicin producer, Micromonospora echinospora ssp. challisensis NRRL 12255 has considerable biosynthetic capabilities as revealed by genome scanning. Among these is a locus containing both type I and type II PKS genes. The presumed products of this locus, TLN-05220 (1) and TLN-05223 (2), bear a core backbone composed of six fused rings starting with a 2-pyridone moiety. The structures were confirmed by conventional spectral analyses including MS, and 1D and 2D NMR experiments. Comparison of both the 1H and 13C NMR data of the newly isolated compound with those of echinosporamicin and bravomicin A led us to propose a revision of the structure of the latter to include a 2-pyridone instead of the pyran originally postulated. Both compounds (1 and 2) possessed strong antibacterial activity against a series of gram-positive pathogens including several strains of methicillin-resistant Staphylococcus aureus and vancomycin-resistant Enterococci (VRE), and cytotoxic activities against several human tumor cell lines. The TLN compounds are the first of this group with reported anticancer activity.

Biosynthesis of diazepinomicin/ECO-4601, a Micromonospora secondary metabolite with a novel ring system.

The novel microbial metabolite diazepinomicin/ECO-4601 (1) has a unique tricyclic dibenzodiazepinone core, which was unprecedented among microbial metabolites. Labeled feeding experiments indicated that the carbocyclic ring and the ring nitrogen of tryptophan could be incorporated via degradation to the 3-hydroxyanthranilic acid, forming ring A and the nonamide nitrogen of 1. Genomic analysis of the biosynthetic locus indicated that the farnesyl side chain was mevalonate derived, the 3-hydroxyanthranilic acid moiety could be formed directly from chorismate, and the third ring was constructed via 3-amino-5-hydroxybenzoic acid. Successful incorporation of 4,6-D2-3-hydroxyanthranilic acid into ring A of 1 via feeding experiments supports the genetic analysis and the allocation of the locus to this biosynthesis. These studies highlight the enzymatic complexity needed to produce this structural type, which is rare in nature.

Reassembly of anthramycin biosynthetic gene cluster by using recombinogenic cassettes.

The reassembly and heterologous expression of complete gene clusters in shuttle vectors has enabled investigations of several large biosynthetic pathways in recent years. With a gene cluster in a mobile construct, the interrogation of gene functions from both culturable and nonculturable organisms is greatly accelerated and large pathway engineering efforts can be executed to produce "new" natural products. However, the genetic manipulation of complete natural product biosynthetic gene clusters is often complicated by their sheer size (10-200 kbp), which makes standard restriction/ligation-based methods impracticable. To circumvent these problems, alternative recombinogenic methods, which depend on engineered homology-based recombination have recently arisen as a powerful alternative. Here, we describe a new general technique that can be used to reconstruct large biosynthetic pathways from overlapping cosmids by retrofitting each cosmid with a "recombinogenic cassette" that contains a shared homologous element and orthogonal antibiotic markers. We employed this technique to reconstruct the anthramycin biosynthetic gene cluster of the thermotolerant actinomycete Streptomyces refuineus, from two >30 kbp cosmids into a single cosmid and integrate it into the genome of Streptomyces lividans. Anthramycin production in the heterologous Streptomyces host confirmed the integrity of the reconstructed pathway and validated the proposed boundaries of the gene cluster. Notably, anthramycin production by recombinant S. lividans was seen only during growth at high temperature--a property also shown by the natural host. This work provides tools to engineer the anthramycin biosynthetic pathway and to explore the connection between anthramycin production and growth at elevated temperatures.

Benzodiazepine biosynthesis in Streptomyces refuineus.

Anthramycin is a benzodiazepine alkaloid with potent antitumor and antibiotic activity produced by the thermophilic actinomycete Streptomyces refuineus sbsp. thermotolerans. In this study, the complete 32.5 kb gene cluster for the biosynthesis of anthramycin was identified by using a genome-scanning approach, and cluster boundaries were estimated via comparative genomics. A lambda-RED-mediated gene-replacement system was developed to provide supporting evidence for critical biosynthetic genes and to validate the boundaries of the proposed anthramycin gene cluster. Sequence analysis reveals that the 25 open reading frame anthramycin cluster contains genes consistent with the biosynthesis of the two halves of anthramycin: 4 methyl-3-hydroxyanthranilic acid and a "dehydroproline acrylamide" moiety. These nonproteinogenic amino acid precursors are condensed by a two-module nonribosomal peptide synthetase (NRPS) terminated by a reductase domain, consistent with the final hemiaminal oxidation state of anthramycin.

Molecular basis for chloronium-mediated meroterpene cyclization: cloning, sequencing, and heterologous expression of the napyradiomycin biosynthetic gene cluster.

Structural inspection of the bacterial meroterpenoid antibiotics belonging to the napyradiomycin family of chlorinated dihydroquinones suggests that the biosynthetic cyclization of their terpenoid subunits is initiated via a chloronium ion. The vanadium-dependent haloperoxidases that catalyze such reactions are distributed in fungi and marine algae and have yet to be characterized from bacteria. The cloning and sequence analysis of the 43-kb napyradiomycin biosynthetic cluster (nap) from Streptomyces aculeolatus NRRL 18422 and from the undescribed marine sediment-derived Streptomyces sp. CNQ-525 revealed 33 open reading frames, three of which putatively encode vanadium-dependent chloroperoxidases. Heterologous expression of the CNQ-525-based nap biosynthetic cluster in Streptomyces albus produced at least seven napyradiomycins, including the new analog 2-deschloro-2-hydroxy-A80915C. These data not only revealed the molecular basis behind the biosynthesis of these novel meroterpenoid natural products but also resulted in the first in vivo verification of vanadium-dependent haloperoxidases.

Genomic analyses lead to novel secondary metabolites. Part 3. ECO-0501, a novel antibacterial of a new class.

Genomic analyses of Amycolatopsis orientalis ATCC 43491 strain, deposited as a vancomycin producer, revealed the presence of genetic loci for the production of at least 10 secondary metabolites other than vancomycin. One of these gene clusters, which contained a type I polyketide synthase, was predicted to direct the synthesis of novel class of compound, a glycosidic polyketide ECO-0501 (1). Screening of culture extracts for a compound with the predicted physicochemical properties of the product from this locus, led to the isolation of the 13-O-glucuronide of 13-hydroxy-2,12,14,16,22-pentamethyl-28-(N-methyl-guanidino)-octacosa-2,4,6,8,10,14,20,24-octaenoic acid (2-hydroxy-5-oxo-cyclopent-1-enyl)-amide (ECO-0501, 1). The structure, confirmed by spectral analyses including MS, and ID and 2D NMR experiments, were in accord with that predicted by genomic analyses. ECO-0501 possessed strong antibacterial activity against a series of Gram-positive pathogens including several strains of methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococci (VRE). ECO-0501 was chemically modified by esterification (1a-1c), N-acetylation (1d) and hydrogenation (1e) in order to explore structure activity relationships (SAR).

Isolation and identification of three new 5-alkenyl-3,3(2H)-furanones from two streptomyces species using a genomic screening approach.

Analyses of biosynthetic gene clusters derived from Streptomyces aculeolatus NRRL 18422 and Streptomyces sp. Eco86 indicated that both microorganisms have similar type I polyketide synthase (PKS) gene clusters with relatively few genes encoding post-PKS elaborative enzymes. However both gene clusters included a sequence coding for a relatively uncommon oxidative enzyme related to Baeyer-Villiger, flavin-type monooxygenases. Screening of culture extracts for compounds with the predicted physicochemical properties of the end products from these loci, led to the isolation of three 5-alkenyl-3,3(2H)-furanones, one (E-837, 1) from the former and two (E-492, 2, E-975, 3) from the latter strain. The structures, confirmed by spectral analyses including MS, and ID and 2D NMR experiments, were in accord with those predicted by genomic analyses. Baeyer-Villiger type oxidation is postulated to be involved in the formation of the furanone moieties in these molecules. All three new compounds were tested for their electron transport inhibitory activities. They had IC50 values of 1-4 microg/ml against Ascaris suum NADH-fumarate reductase and 1-12 microg/ml against bovine heart NADH oxidase.

Microbial genomics as a guide to drug discovery and structural elucidation: ECO-02301, a novel antifungal agent, as an example.

Analysis of the genome of Streptomyces aizunensis NRRL B-11277 indicated its potential to produce a compound of novel and highly predictable structure. The structure was predicted with sufficient accuracy to allow straightforward detection of the specific metabolite in HPLC profiles of fermentation extracts and hence to guide the isolation. The spectroscopic work was reduced to a confirmation of structure rather than a first principle determination. The compound, ECO-02301 (1), demonstrated potent antifungal activity. This work exemplifies not only the discovery of novel antibiotics from well-characterized organisms but also the utility of genomics as a further tool, complementary to spectroscopy, to enable rapid determination of complex structures.

The neocarzinostatin biosynthetic gene cluster from Streptomyces carzinostaticus ATCC 15944 involving two iterative type I polyketide synthases.

The biosynthetic gene cluster for the enediyne antitumor antibiotic neocarzinostatin (NCS) was localized to 130 kb continuous DNA from Streptomyces carzinostaticus ATCC15944 and confirmed by gene inactivation. DNA sequence analysis of 92 kb of the cloned region revealed 68 open reading frames (ORFs), 47 of which were determined to constitute the NCS cluster. Sequence analysis of the genes within the NCS cluster suggested dNDP-D-mannose as a precursor for the deoxy aminosugar, revealed two distinct type I polyketide synthases (PKSs), and supported a convergent model for NCS chromophore biosynthesis from the deoxy aminosugar, naphthoic acid, and enediyne core building blocks. These findings shed light into deoxysugar biosynthesis, further support the iterative type I PKS paradigm for enediyne core biosynthesis, and unveil a mechanism for microbial polycyclic aromatic polyketide biosynthesis by an iterative type I PKS.

A genomics-guided approach for discovering and expressing cryptic metabolic pathways.

Genome analysis of actinomycetes has revealed the presence of numerous cryptic gene clusters encoding putative natural products. These loci remain dormant until appropriate chemical or physical signals induce their expression. Here we demonstrate the use of a high-throughput genome scanning method to detect and analyze gene clusters involved in natural-product biosynthesis. This method was applied to uncover biosynthetic pathways encoding enediyne antitumor antibiotics in a variety of actinomycetes. Comparative analysis of five biosynthetic loci representative of the major structural classes of enediynes reveals the presence of a conserved cassette of five genes that includes a novel family of polyketide synthase (PKS). The enediyne PKS (PKSE) is proposed to be involved in the formation of the highly reactive chromophore ring structure (or "warhead") found in all enediynes. Genome scanning analysis indicates that the enediyne warhead cassette is widely dispersed among actinomycetes. We show that selective growth conditions can induce the expression of these loci, suggesting that the range of enediyne natural products may be much greater than previously thought. This technology can be used to increase the scope and diversity of natural-product discovery.

The calicheamicin gene cluster and its iterative type I enediyne PKS.

The enediynes exemplify nature's ingenuity. We have cloned and characterized the biosynthetic locus coding for perhaps the most notorious member of the nonchromoprotein enediyne family, calicheamicin. This gene cluster contains an unusual polyketide synthase (PKS) that is demonstrated to be essential for enediyne biosynthesis. Comparison of the calicheamicin locus with the locus encoding the chromoprotein enediyne C-1027 reveals that the enediyne PKS is highly conserved among these distinct enediyne families. Contrary to previous hypotheses, this suggests that the chromoprotein and nonchromoprotein enediynes are generated by similar biosynthetic pathways.