Search

Phenazine-type metabolites arise from either phenazine-1-carboxylic acid (PCA) or phenazine-1,6-dicarboxylic acid (PDC). Although the biosynthesis of PCA has been studied extensively, PDC assembly remains unclear. Esmeraldins and saphenamycin, the PDC originated products, are antimicrobial and antitumor metabolites isolated from Streptomyces antibioticus Tü 2706. Herein, the esmeraldin biosynthetic gene cluster was identified on a dispensable giant plasmid. Twenty-four putative esm genes were characterized by bioinformatics, mutagenesis, genetic complementation, and functional protein expressions. Unlike enzymes involved in PCA biosynthesis, EsmA1 and EsmA2 together decisively promoted the PDC yield. The resulting PDC underwent a series of conversions to give 6-acetylphenazine-1-carboxylic acid, saphenic acid, and saphenamycin through a unique one-carbon extension by EsmB1-B5, a keto reduction by EsmC, and an esterification by EsmD1-D3, the atypical polyketide sythases, respectively. Two transcriptional regulators, EsmT1 and EsmT2, are required for esmeraldin production., (Copyright © 2012 Elsevier Ltd. All rights reserved.)

The amide synthase of the geldanamycin producer, Streptomyces hygroscopicus, shows a broader chemoselectivity than the corresponding amide synthase present in Actinosynnema pretiosum, the producer of the highly cytotoxic ansamycin antibiotics, the ansamitocins. This was demonstrated when blocked mutants of both strains incapable of biosynthesizing 3-amino-5-hydroxybenzoic acid (AHBA), the polyketide synthase starter unit of both natural products, were supplemented with 3-amino-5-hydroxymethylbenzoic acid instead. Unlike the ansamitocin producer A. pretiosum, S. hygroscopicus processed this modified starter unit not only to the expected 19-membered macrolactams but also to ring enlarged 20-membered macrolactones. The former mutaproducts revealed the sequence of transformations catalyzed by the post-PKS tailoring enzymes in geldanamycin biosynthesis. The unprecedented formation of the macrolactones together with molecular modeling studies shed light on the mode of action of the amide synthase responsible for macrocyclization. Obviously, the 3-hydroxymethyl substituent shows similar reactivity and accessibility toward C-1 of the seco-acid as the arylamino group, while phenolic hydroxyl groups lack this propensity to act as nucleophiles in the macrocyclization. The promiscuity of the amide synthase of S. hygroscopicus was further demonstrated by successful feeding of four other m-hydroxymethylbenzoic acids, leading to formation of the expected 20-membered macrocycles. Good to moderate antiproliferative activities were encountered for three of the five new geldanamycin derivatives, which matched well with a competition assay for Hsp90α., (© 2011 American Chemical Society)

Ansamitocins are potent antitumor agents produced by Actinosynnema pretiosum. As deduced from their structures, an N-methylation on the amide bond is required among the various modifications. The protein encoded by asm10 belongs to the SAM-dependent methyltransferase family. Through gene inactivation and complementation, asm10 was proved to be responsible for the N-methylation of ansamitocins. Asm10 is a 33.0 kDa monomer, as determined by gel filtration. By using N-desmethyl-ansamitocin P-3 as substrate, the optimal temperature and pH for Asm10 catalysis were determined to be 32 °C and 10.0, respectively. Asm10 also showed broad substrate flexibility toward other N-desmethyl-ansamycins and synthetic indolin-2-ones. Through site-directed mutagenesis, Asp154 and Leu155 of Asm10 were confirmed to be essential for its catalysis, possibly through the binding of SAM. The characterization of this unique N-methyltransferase has enriched the toolbox for engineering N-methylated derivatives from both natural and synthetic compounds; this will allow known potential drugs to be modified., (Copyright © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.)

The aminoshikimate pathway of formation of 3-amino-5-hydroxybenzoic acid (AHBA), the precursor of ansamycin and other antibiotics is reviewed. In this biosynthesis, genes for kanosamine formation have been recruited from other genomes, to provide a nitrogenous precursor. Kanosamine is then phosphorylated and converted by common cellular enzymes into 1-deoxy-1-imino-erythrose 4-phosphate, the substrate for the formation of aminoDAHP. This is converted via 5-deoxy-5-aminodehydroquinic acid and 5-deoxy-5-aminodehydroshikimic acid into AHBA. Remarkably, the pyridoxal phosphate enzyme AHBA synthase seems to have two catalytic functions: As a homodimer, it catalyzes the last reaction in the pathway, the aromatization of 5-deoxy-5-aminodehydroshikimic acid, and at the beginning of the pathway in a complex with the oxidoreductase RifL it catalyzes the transamination of UDP-3-keto-D-glucose. The AHBA synthase gene also serves as a useful tool in the genetic screening for new ansamycins and other AHBA-derived natural products.

Asukamycin, a member of the manumycin family metabolites, is an antimicrobial and potential antitumor agent isolated from Streptomyces nodosus subsp. asukaensis. The entire asukamycin biosynthetic gene cluster was cloned, assembled, and expressed heterologously in Streptomyces lividans. Bioinformatic analysis and mutagenesis studies elucidated the biosynthetic pathway at the genetic and biochemical level. Four gene sets, asuA-D, govern the formation and assembly of the asukamycin building blocks: a 3-amino-4-hydroxybenzoic acid core component, a cyclohexane ring, two triene polyketide chains, and a 2-amino-3-hydroxycyclopent-2-enone moiety to form the intermediate protoasukamycin. AsuE1 and AsuE2 catalyze the conversion of protoasukamycin to 4-hydroxyprotoasukamycin, which is epoxidized at C5-C6 by AsuE3 to the final product, asukamycin. Branched acyl CoA starter units, derived from Val, Leu, and Ile, can be incorporated by the actions of the polyketide synthase III (KSIII) AsuC3/C4 as well as the cellular fatty acid synthase FabH to produce the asukamycin congeners A2-A7. In addition, the type II thioesterase AsuC15 limits the cellular level of omega-cyclohexyl fatty acids and likely maintains homeostasis of the cellular membrane.

The timing of introduction of the unusually placed Delta(11,13) diene system in ansamitocin (AP) biosynthesis was probed by synthesizing optically active potential tri- and tetraketide intermediates as their SNAC thioesters. An AP-nonproducing mutant Actinosynnema pretiosum was complemented by the R enantiomer of the triketide and by the tetraketide with rearranged double bonds, but not by the tetraketide carrying the double bonds in conjugation to the thioester function. The results show that the double bonds are installed in their final positions during processing of the nascent polyketide on module 3 of the asm PKS and that KS4 of the PKS acts as a gatekeeper which accepts only a tetraketide with shifted double bonds as substrate for further processing.

Ansamitocins are potent antitumor maytansinoids produced by Actinosynnema pretiosum. Their biosynthesis involves the initial assembly of a macrolactam polyketide, followed by a series of postpolyketide synthase (PKS) modifications. Three ansamitocin glycosides were isolated from A. pretiosum and fully characterized structurally as novel ansamitocin derivatives, carrying a beta-D-glucosyl group attached to the macrolactam amide nitrogen in place of the N-methyl group. By gene inactivation and complementation, asm25 was identified as the N-glycosyltransferase gene responsible for the macrolactam amide N-glycosylation of ansamitocins. Soluble, enzymatically active Asm25 protein was obtained from asm25-expressing E. coli by solubilization from inclusion bodies. Its optimal reaction conditions, including temperature, pH, metal ion requirement, and Km/Kcat, were determined. Asm25 also showed broad substrate specificity toward other ansamycins and synthetic indolin-2-ones. To the best of our knowledge, this represents the first in vitro characterization of a purified antibiotic N-glycosyltransferase.

Feeding experiments with isotope-labeled precursors rule out hydroxypyruvate and TCA cycle intermediates as the metabolic source of methoxymalonyl-ACP, the substrate for incorporation of "glycolate" units into ansamitocin P-3, soraphen A, and other antibiotics. They point to 1,3-bisphosphoglycerate as the source of the methoxymalonyl moiety and show that its C-1 gives rise to the thioester carbonyl group (and hence C-1 of the "glycolate" unit), and its C-3 becomes the free carboxyl group of methoxymalonyl-ACP, which is lost in the subsequent Claisen condensation on the type I modular polyketide synthases (PKS). d-[1,2-(13)C(2)]Glycerate is also incorporated specifically into the "glycolate" units of soraphen A, but not of ansamitocin P-3, suggesting differences in the ability of the producing organisms to activate glycerate. A biosynthetic pathway from 1,3-bisphosphoglycerate to methoxymalonyl-ACP is proposed. Two new syntheses of R- and S-[1,2-(13)C(2)]glycerol were developed as part of this work.

The biosynthesis of the antitumor antibiotic, ansamitocin, involves the assembly of a linear octaketide on the ansamitocin (asm) polyketide synthase (PKS), which is then cyclized to proansamitocin and further modified to the final product. In the first chain-extension step on the asm PKS, a stereocenter is generated which is then obliterated in a subsequent double-bond migration. The cryptic configuration at this stereocenter was determined by first synthesizing the two enantiomers of the intermediate diketide as their N-acetylcysteamine (SNAC) thioesters. These were then used to demonstrate that only the R enantiomer complements a 3-amino-5-hydroxybenzoic acid (AHBA) deficient mutant of Actinosynnema pretiosum to restore ansamitocin formation. The low efficiency of complementation by the diketide, compared to AHBA, is due to inefficient loading onto the PKS and not the inhibition of the enzyme. A presumed next chain-extension intermediate-the triketide with an unrearranged double bond-was also synthesized as its SNAC ester, but did not complement the AHBA(-) mutant.

Because of their ecological functions, natural products have been optimized in evolution for interaction with biological systems and receptors. However, they have not necessarily been optimized for other desirable drug properties and thus can often be improved by structural modification. Using examples from the literature, this paper reviews the opportunities for increasing structural diversity among natural products by combinatorial biosynthesis, i.e., the genetic manipulation of biosynthetic pathways. It distinguishes between combinatorial biosynthesis in a narrower sense to generate libraries of modified structures, and metabolic engineering for the targeted formation of specific structural analogs. Some of the problems and limitations encountered with these approaches are also discussed. Work from the author's laboratory on ansamycin antibiotics is presented which illustrates some of the opportunities and limitations.

A 45 kb DNA sequencing analysis from Streptomyces hygroscopicus 5008 involved in validamycin A (VAL-A) biosynthesis revealed 16 structural genes, 2 regulatory genes, 5 genes related transport, transposition/integration or tellurium resistance; another 4 genes had no obvious identity. The VAL-A biosynthetic pathway was proposed, with assignment of the required genetic functions confined to the sequenced region. A cluster of eight reassembled genes was found to support VAL-A synthesis in a heterologous host, S. lividans 1326. In vivo inactivation of the putative glycosyltransferase gene (valG) abolished the final attachment of glucose for VAL production and resulted in accumulation of the VAL-A precursor, validoxylamine, while the normal production of VAL-A could be restored by complementation with valG. The role of valG in the glycosylation of validoxylamine to VAL-A was demonstrated in vitro by enzymatic assay.

In this review the author traces his scientific career from its beginnings in Germany to his moves to, successively, Purdue University, The Ohio State University, and finally University of Washington. During this time his research progressed from extensive studies on ergot alkaloids, the stereochemistry of enzyme reactions, and tracer studies on antibiotic biosynthesis to its latest emphasis on the molecular biology of ansamycin antibiotics. The formative influence of several mentors and colleagues is acknowledged.

A gene cluster responsible for the biosynthesis of validamycin, an aminocyclitol antibiotic widely used as a control agent for sheath blight disease of rice plants, was identified from Streptomyces hygroscopicus subsp. jinggangensis 5008 using heterologous probe acbC, a gene involved in the cyclization of D-sedoheptulose 7-phosphate to 2-epi-5-epi-valiolone of the acarbose biosynthetic gene cluster originated from Actinoplanes sp. strain SE50/110. Deletion of a 30-kb DNA fragment from this cluster in the chromosome resulted in loss of validamycin production, confirming a direct involvement of the gene cluster in the biosynthesis of this important plant protectant. A sequenced 6-kb fragment contained valA (an acbC homologue encoding a putative cyclase) as well as two additional complete open reading frames (valB and valC, encoding a putative adenyltransferase and a kinase, respectively), which are organized as an operon. The function of ValA was genetically demonstrated to be essential for validamycin production and biochemically shown to be responsible specifically for the cyclization of D-sedoheptulose 7-phosphate to 2-epi-5-epi-valiolone in vitro using the ValA protein heterologously overexpressed in E. coli. The information obtained should pave the way for further detailed analysis of the complete biosynthetic pathway, which would lead to a complete understanding of validamycin biosynthesis.

Rifamycin B biosynthesis by Amycolatopsis mediterranei S699 involves a number of unusual modification reactions in the formation of the unique polyketide backbone and decoration of the molecule. A number of genes believed to be involved in the tailoring of rifamycin B were investigated and the results confirmed that the formation of the naphthalene ring moiety of rifamycin takes place during the polyketide chain extension and is catalysed by Rif-Orf19, a 3-(3-hydroxyphenyl)propionate hydroxylase-like protein. The cytochrome P450-dependent monooxygenase encoded by rif-orf5 is required for the conversion of the Delta12, 29 olefinic bond in the polyketide backbone of rifamycin W into the ketal moiety of rifamycin B. Furthermore, Rif-Orf3 may be involved in the regulation of rifamycin B production, as its knock-out mutant produced about 40 % more rifamycin B than the wild-type. The work also revealed that many of the genes located in the cluster are not involved in rifamycin biosynthesis, but might be evolutionary remnants carried over from an ancestral lineage.

Maytansine, a potent clinically evaluated plant-derived anti-tumor drug, and its microbial counterpart, ansamitocin P-3, showed a substantially higher cytoxicity than many other anti-tumor drugs. Owing to a shortage of material and lack of sufficiently sensitive analytical methods at the time, no metabolism studies were apparently carried out in conjunction with the initial preclinical and clinical studies on maytansine, but some products of decomposition during the period of storage of the formulated drug were reported. In the current study, the in vitro metabolism of maytansine and ansamitocin P-3 was studied after incubation with rat and human liver microsomes in the presence of NADPH, and with rat and human plasma and whole blood, using liquid chromatography/multi-stage mass spectrometry. Unchanged ansamitocin P-3 and 11 metabolites and unchanged maytansine and seven metabolites were profiled and the structures of some metabolites were tentatively assigned based on their multi-stage electrospray ion-trap mass fragmentation data and in some cases accurate mass measurement. The major pathway of ansamitocin P-3 metabolism in human liver microsomes appears to be demethylation at C-10. Oxidation and sequential oxidation/demethylation also occurred, although to a lesser extent. However, the major pathway of maytansine metabolism in human liver microsomes is N-demethylation of the methylamide of the ester moiety. Several minor pathways including O/N-demethylation, oxidation and hydrolysis of the ester bond were also observed. There were no differences in maytansine metabolism between rat and human liver microsomes; however, the rate of metabolism of ansamitocin P-3 was different in rat and human liver microsomes. About 20% of ansamitocin P-3 was converted to its metabolites in rat liver microsomes and about 70% in human liver microsomes under the same conditions. Additionally, 10-O-demethylated ansamitocin P-3 was also detected in the urine after i.v. bolus administration of ansamitocin P-3 to Sprague-Dawley male rats. No metabolites were detected following incubation of maytansine and ansamitocin P-3 with human and rat whole blood and plasma., (Copyright (c) 2005 John Wiley & Sons, Ltd.)

Ansamitocin P-3 (AP3) is a potent maytansinoid antitumor agent isolated from microorganisms and mosses. In this study, a highly sensitive and specific electrospray ionization (ESI) liquid chromatography-tandem mass spectrometry (LC/MS/MS) method for quantitation of AP3 was developed and validated. AP3 was extracted from rat plasma along with the internal standard, depsipeptide FK228 (NSC-630176, FR) with ethyl acetate. Components in the extract were separated on a 50mm x 2.1mm Betabasic C 85 microm stainless steel column by isocratic elution with 70% acetonitrile/0.9% formic acid. The liquid flow was passed through a pre-source splitter and 5% of the eluent was introduced into the API source. The components were analyzed in the multiple-reaction-monitoring (MRM) mode as the precursor/product ion pair of m/z 635.2/547.2 for AP3 and of m/z 541.5/424.0 for the internal standard FR. Linear calibration curves were obtained in the range 1-500 ng/mL using 0.2 mL rat plasma. The within-day coefficients of variation (CVs) were 12.9, 6.7, and 5.5% and the between-day CVs were 10.4, 6.5, and 6.4% (all n = 5) at 1, 10, and 200 ng/mL, respectively. A formulation based on normal saline and PEG300 was then developed and Sprague-Dawley male rats were given this formulated drug by i.v. bolus. Plasma drug concentrations were measured by this method and the pharmacokinetics were analyzed by standard techniques. Plasma concentration-time profiles were found to follow a triexponential decline and the terminal phase was nearly flat, suggesting that the drug distributed in deep tissue compartments or organs and then equilibrates slowly with the blood stream.

Asukamycin (2), a metabolite of Streptomyces nodosus ssp. asukaensis ATCC 29757 and a member of the manumycin family of antibiotics, is assembled from three components, an "upper" polyketide chain initiated by cyclohexanecarboxylic acid, a "lower" polyketide chain initiated by the novel starter unit, 3-amino-4-hydroxybenzoic acid (3,4-AHBA), and a cyclized 5-aminolevulinic acid moiety, 2-amino-3-hydroxycyclopent-2-enone (C(5)N unit). To shed light on the order in which these components are assembled, we synthesized in labeled form various potential intermediates and evaluated their incorporation into 2. The assembly of the molecular framework of 2 from 3,4-AHBA and cyclohexanecarboxylic acid apparently does not involve free, unactivated intermediates. However, protoasukamycin (12), the total synthesis of which is reported, was efficiently converted into 2, demonstrating that the modification of the aromatic ring to the epoxyquinol structure is the terminal step in the biosynthesis. The results suggest that the two polyketide chains are synthesized separately and that the "upper" chain must be connected to the "lower" polyketide chain before the C(5)N unit.

Maytansine and its congeners have been isolated from higher plants, mosses and from an Actinomycete, Actinosynnema pretiosum. Many of these compounds are antitumor agents of extraordinary potency, yet phase II clinical trials with maytansine proved disappointing. The chemistry and biology of maytansinoids has been reviewed repeatedly in the late 1970s and early 1980s; the present review covers new developments in this field during the last two decades. These include the use of maytansinoids as "warheads" in tumor-specific antibodies, preliminary metabolism studies, investigations of their biosynthesis at the biochemical and genetic level, and ecological issues related to the occurrence of such typical microbial metabolites in higher plants.

The functions of six genes in the ansamitocin biosynthetic gene cluster of Actinosynnema pretiosum have been investigated by gene inactivation and chemical analysis of the mutants. They encode a halogenase (asm12), a carbamoyltransferase (asm21), a 20-O-methyltransferase (asm7), a 3-O-acyltransferase (asm19), an epoxidase (asm11), and an N-methyltransferase (asm10), respectively, and are responsible for the six post-PKS modification steps in ansamitocin formation. Several of the enzymes have relaxed substrate specificities, resulting in multiple parallel pathways in a metabolic grid, albeit with a preferred sequence of reactions as listed above.

The gene rif orf14 in the rifamycin biosynthetic gene cluster of Amycolatopsis mediterranei S699, producer of the antitubercular drug rifamycin B, encodes a protein of 272 amino acids identified as an AdoMet: 27-O-demethylrifamycin SV methyltransferase. Frameshift inactivation of rif orf14 generated a mutant of A. mediterranei S699 that produces no rifamycin B, but accumulates 27-O-demethylrifamycin SV (DMRSV) as the major new metabolite, together with a small quantity of 27-O-demethyl-25-O-desacetylrifamycin SV (DMDARSV). Heterologous expression of rif orf14 in Escherichia coli yielded a 33.8-kDa polyhistidine-tagged polypeptide, which efficiently catalyzes the methylation of DMRSV to rifamycin SV, but not that of DMDARSV or rifamycin W. 27-O-Demethylrifamycin S was methylated poorly, if at all, by the enzyme to produce rifamycin S. The purified enzyme does not require a divalent cation for catalytic activity. While Ca(2+) or Mg(2+) inhibits the enzyme activity slightly, Zn(2+), Ni(2+), and Co(2+) are strongly inhibitory. The K(m) values for DMRSV and S-adenosyl-L-methionine (AdoMet) are 18.0 and 19.3 microM, respectively, and the K(cat) is 87s(-1). The results indicate that DMRSV is a direct precursor of rifamycin SV and that acetylation of the C-25 hydroxyl group must precede the methylation reaction. They also suggest that rifamycin S is not the precursor of rifamycin SV in rifamycin B biosynthesis, but rather an oxidative shunt-product.

Individual plants belonging to different species of the family Celastraceae collected from their natural habitats in South Africa (Putterlickia verrucosa (E. Meyer ex Sonder) Szyszyl., Putterlickia pyracantha (L.) Szyszyl., Putterlickia retrospinosa van Wyk and Mostert) and Brazil (Maytenus ilicifolia Mart. ex Reiss., Maytenus evonymoides Reiss., Maytenus aquifolia Mart.) were investigated for the presence of maytansinoids and of maytansine, an ansamycin of high cytotoxic activity. Maytansinoids were not detectable in plants grown in Brazil. Analysis of plants growing in South Africa, however, showed clearly that maytansinoids were present in some individual plants but were not detectable in others. Molecular biological analysis of a Putterlickia verrucosa cell culture gave no evidence for the presence of the aminohydroxybenzoate synthase gene which is unique to the biosynthesis of aminohydroxybenzoate, a precursor of the ansamycins including maytansinoids. Moreover, this gene was not detectable in DNA extracted from the aerial parts of Putterlickia plants. In contrast, observations indicate that this gene may be present in microbes of the rhizosphere of Putterlickia plants. Our observations are discussed with respect to the possibility that the roots of Putterlickia plants may be associated with microorganisms which are responsible for the biosynthesis of maytansine or maytansinoids.

The biosynthesis of 3-amino-5-hydroxybenzoic acid (AHBA), precursor of the ansamycin and mitomycin antibiotics, proceeds by the aminoshikimate pathway from 3,4-dideoxy-4-amino-D-arabino-heptulosonic acid 7-phosphate (aminoDAHP). Identification of RifN, product of one of three genes from the rifamycin biosynthetic gene cluster known to be essential for aminoDAHP formation, as a specific kanosamine (3-deoxy-3-amino-D-glucose) 6-kinase establishes kanosamine and its 6-phosphate as specific intermediates in AHBA formation. This suggests a hypothetical reaction sequence for aminoDAHP formation, and thus for the early steps of AHBA biosynthesis, starting from UDP-D-glucose and introducing the nitrogen by oxidation and transamination at C-3.

The potent antitumor activity of the ansamitocins, polyketides isolated from Actinosynnema pretiosum, is absolutely dependent on a short acyl group esterified to the C-3 oxygen of the macrolactam ring. Asm19, a gene in the ansamitocin biosynthetic gene cluster with homology to macrolide O-acyltransferase genes, is thought to encode the enzyme catalyzing this esterification. A mutant carrying an inactivated asm19 no longer produced ansamitocins but accumulated N-desmethyl-4,5-desepoxymaytansinol, rather than maytansinol, indicating that the acylation is not the terminal step of the biosynthetic sequence. Bioconversion experiments and in vitro studies with recombinant Asm19, expressed in Escherichia coli, showed that the enzyme is very specific toward its alcohol substrate, converting N-desmethyl-4,5-desepoxymaytansinol (but not maytansinol) into ansamitocins, but rather promiscuous toward its acyl substrate, utilizing acetyl-, propionyl-, butyryl-, isobutyryl-, as well as isovaleryl-CoA.

Maytansinoids are potent antitumor agents found in plants and microorganisms. To elucidate their biosynthesis at the biochemical and genetic level and to set the stage for their structure modification through genetic engineering, we have cloned two gene clusters required for the biosynthesis of the maytansinoid, ansamitocin, from a cosmid library of Actinosynnema pretiosum ssp. auranticum ATCC 31565. This is a rare case in which the genes involved in the formation of a secondary metabolite are dispersed in separate regions in an Actinomycete. A set of genes, asm22-24, asm43-45, and asm47, was identified for the biosynthesis of the starter unit, 3-amino-5-hydroxybenzoic acid (AHBA). Remarkably, there are two AHBA synthase gene homologues, which may have different functions in AHBA formation. Four type I polyketide synthase genes, asmA-D, followed by the downloading asm9, together encode eight homologous sets of enzyme activities (modules), each catalyzing a specific round of chain initiation, elongation, or termination steps, which assemble the ansamitocin polyketide backbone. Another set of genes, asm13-17, encodes the formation of an unusual "methoxymalonate" polyketide chain extension unit that, notably, seems to be synthesized on a dedicated acyl carrier protein rather than as a CoA thioester. Additional ORFs are involved in postsynthetic modifications of the initial polyketide synthase product, which include methylations, an epoxidation, an aromatic chlorination, and the introduction of acyl and carbamoyl groups. Tentative functions of several asm genes were confirmed by inactivation and heterologous expression.

A subcluster of five genes, asm13-17, from the ansamitocin biosynthetic gene cluster of Actinosynnema pretiosum was coexpressed in Streptomyces lividans with the genes encoding the 6-deoxyerythronolide B (6-DEB) synthase from Saccharopolyspora erythraea, in which the methylmalonate-specifying AT6 domain had been replaced by the methoxymalonate-specifying AT8 domain from the FK520 cluster of Streptomyces hygroscopicus. The engineered strain produced the predicted product, 2-desmethyl-2-methoxy-DEB, instead of 6-DEB and 2-desmethyl-6-DEB, which were formed in the absence of the asm13-17 cassette, indicating that asm13-17 are sufficient for synthesis of this unusual chain extension unit. Deletion of asm17, encoding a methyltransferase, from the cassette gave 6-DEB instead of its hydroxy analogue, indicating that methylation of the extender unit is required for its incorporation.

The unusual "glycolate" extender unit at C-9/C-10 of ansamitocin is not derived from 2-hydroxymalonyl-CoA or 2-methoxymalonyl-CoA, as demonstrated by feeding experiments with the corresponding 1-13C-labeled N-acetylcysteamine thioesters but is formed from an acyl carrier protein (ACP)-bound substrate, possibly 2-methoxymalonyl-ACP, elaborated by enzymes encoded by a subcluster of five genes, asm12-17, from the ansamitocin bisosynthetic gene cluster.

To study the biosynthesis of the pseudotetrasaccharide acarbose, dTDP-4-amino-4,6-dideoxy-alpha-D-glucose (3) was prepared from galactose in 16 steps. After initial protecting-group manipulations, the 6-position of galactose was deoxygenated by hydride displacement of a tosylate. Similarly a tosyl group at the 4-position was displaced upon reaction with sodium azide. Conversion of the resulting glycoside to a glycosyl phosphate was accomplished by reaction of a glycosyl trichloroacetimidate with dibenzyl phosphate. Subsequent removal of the benzyl protecting groups and reduction of the azide by hydrogenation and coupling with an activated nucleoside phosphate gave dTDP-4-amino-4,6-dideoxy-alpha-D-glucose.

The evolution of the field of biosynthesis from the unravelling of the mode of formation of natural products to the use of such knowledge to create new compounds is reviewed using examples from the author's laboratory. The discussion focuses on the mode of operation of type II (spore pigment PKS) and type I (rifamycin PKS) polyketide synthases and their diversion to generate unnatural products, and on the genetics and biochemistry of deoxysugar formation in granaticin biosynthesis as a prerequisite to combinatorial enzymatic synthesis of unusual glycosides.

To investigate a novel branch of the shikimate biosynthesis pathway operating in the formation of 3-amino-5-hydroxybenzoic acid (AHBA), the unique biosynthetic precursor of rifamycin and related ansamycins, a series of target-directed mutations and heterologous gene expressions were investigated in Amycolatopsis mediterranei and Streptomyces coelicolor. The genes involved in AHBA formation were inactivated individually, and the resulting mutants were further examined by incubating the cell-free extracts with known intermediates of the pathway and analyzing for AHBA formation. The rifL, -M, and -N genes were shown to be involved in the step(s) from either phosphoenolpyruvate/d-erythrose 4-phosphate or other precursors to 3,4-dideoxy-4-amino-d-arabino-heptulosonate 7-phosphate. The gene products of the rifH, -G, and -J genes resemble enzymes involved in the shikimate biosynthesis pathway (August, P. R., Tang, L., Yoon, Y. J., Ning, S., Müller, R., Yu, T.-W., Taylor, M., Hoffmann, D., Kim, C.-G., Zhang, X., Hutchinson, C. R., and Floss, H. G. (1998) Chem. Biol. 5, 69-79). Mutants of the rifH and -J genes produced rifamycin B at 1% and 10%, respectively, of the yields of the wild type; inactivation of the rifG gene did not affect rifamycin production significantly. Finally, coexpressing the rifG-N and -J genes in S. coelicolor YU105 under the control of the act promoter led to significant production of AHBA in the fermented cultures, confirming that seven of these genes are indeed necessary and sufficient for AHBA formation. The effects of deletion of individual genes from the heterologous expression cassette on AHBA formation duplicated the effects of the genomic rifG-N and -J mutations on rifamycin production, indicating that all these genes encode proteins with catalytic rather than regulatory functions in AHBA formation for rifamycin biosynthesis by A. mediterranei.

Five new type II manumycins, containing the hydroxyquinol mC7N unit, asukamycins A-II, B-II, C-II, D-II, E-II, were discovered in cultures of Streptomyces nodosus ssp. asukaensis. The biosynthetic origin of the type II manumycins from the type I compounds, containing an epoxyquinol mC7N unit, was deduced from the time course of production and proven by preparing [7'-13C]asukamycin A and demonstrating its incorporation into asukamycin A-II.

To study the biosynthesis of the pseudotrisaccharide antibiotic, validamycin A (1), a number of potential precursors of the antibiotic were synthesized in (2)H-, (3)H-, or (13)C-labeled form and fed to cultures of Streptomyces hygroscopicus var. limoneus. The resulting validamycin A from each of these feeding experiments was isolated, purified and analyzed by liquid scintillation counting, (2)H- or (13)C NMR or selective ion monitoring mass spectrometry (SIM-MS) techniques. The results demonstrate that 2-epi-5-epi-valiolone (9) is specifically incorporated into 1 and labels both cyclitol moieties. This suggests that 9 is the initial cyclization product generated from an open-chain C(7) precursor, D-sedoheptulose 7-phosphate (5), by a DHQ synthase-like cyclization mechanism. A more proximate precursor of 1 is valienone (11), which is also incorporated into both cyclitol moieties. The conversion of 9 into 11 involves first epimerization to 5-epi-valiolone (10), which is efficiently incorporated into 1, followed by dehydration, although a low level of incorporation of 2-epi-valienone (15) is also observed. Reduction of 11 affords validone (12), which is also incorporated specifically into 1, but labels only the reduced cyclitol moiety. The mode of introduction of the nitrogen atom linking the two pseudosaccharide moieties is not clear yet. 7-Tritiated valiolamine (8), valienamine (2), and validamine (3) were all not incorporated into 1, although each of these amines has been isolated from the fermentation, with 3 being most prevalent. Demonstration of in vivo formation of [7-(3)H]validamine ([7-(3)H]-3) from [7-(3)H]-12 suggests that 3 may be a pathway intermediate and that the nonincorporation of [7-(3)H]-3 into 1 is due to a lack of cellular uptake. We thus propose that 3, formed by amination of 12, and 11 condense to form a Schiff base, which is reduced to the pseudodisaccharide unit, validoxylamine A (13). Transfer of a D-glucose unit to the 4'-position of 13 then completes the biosynthesis of 1. Other possibilities for the mechanism of formation of the nitrogen bridge between the two pseudosaccharide units are also discussed.

Four genes, ORF 22 approximately 25, from the granaticin biosynthetic gene cluster of Streptomyces violaceoruber Tü22 were analyzed for their involvement in the biosynthesis of the two deoxysugar moieties of the granaticins. Each gene was individually inactivated on a cosmid carrying the entire gra gene cluster and the mutant cosmids were transformed into S. coelicolor CH999. Analysis of the pattern of pigment production by the transformants revealed that each of the four ORFs is required for the formation/attachment of the L-rhodinose moiety of granaticin B, but not that of the D-olivose moiety of granaticin. Based on these results and sequence homologies a pathway of dTDP-L-rhodinose formation is proposed which implicates ORF23, and possibly also ORF 24, in the 3-deoxygenation reaction, ORF 25 in the epimerization and ORF 22 in the final 4-ketoreduction.

The studies reported here have established the biosynthetic origin of the mC7N units of acarbose and validamycin from sedo-heptulose 7-phosphate, and have identified 2-epi-5-epi-valiolone as the initial cyclization product. The deoxyhexose moiety of acarbose arises from glucose with deoxythymidyl-diphospho-4-keto-6-deoxy-D-glucose (dTDP-4-keto-6-deoxy-D-glucose) as a proximate intermediate. However, despite the identical origin of the aminocyclitol moieties in acarbose and validamycin A, the pathways of their formation seem to be substantially different. Validamycin A formation involves a number of discrete ketocyclitol intermediates, 5-epi-valiolone, valienone, and validone, whereas no free intermediates have been identified on the pathway from 2-epi-5-epi-valiolone to the pseudodisaccharide moiety of acarbose. The stage is now set for unraveling the mechanism or mechanisms by which the two components of the pseudodisaccharide moieties of acarbose and validamycin are uniquely coupled to each other via a nitrogen bridge.

Background: The committed step in the biosynthesis of the anticancer drug taxol in yew (Taxus) species is the cyclization of geranylgeranyl diphosphate to taxa-4(5),11(12)-diene. The enzyme taxadiene synthase catalyzes this complex olefin cation cyclization cascade involving the formation of three rings and three stereogenic centers., Results: Recombinant taxadiene synthase was incubated with specifically deuterated substrates, and the mechanism of cyclization was probed using MS and NMR analyses of the products to define the crucial hydrogen migration and terminating deprotonation steps. The electrophilic cyclization involves the ionization of the diphosphate with closure of the A-ring, followed by a unique intramolecular transfer of the C11 proton to the re-face of C7 to promote closure of the B/C-ring juncture, and cascade termination by proton elimination from the beta-face of C5., Conclusions: These findings provide insight into the molecular architecture of the first dedicated step of taxol biosynthesis that creates the taxane carbon skeleton, and they have broad implications for the general mechanistic capability of the large family of terpenoid cyclization enzymes.

The role of two thioesterase genes in the premature release of polyketide synthase intermediates during rifamycin biosynthesis in the Amycolatopsis mediterranei S699 strain was investigated. Creation of an in-frame deletion in the rifR gene led to a 30 approximately 60% decrease in the production of both rifamycin B by the S699 strain or a series of tetra- to decaketide shunt products of polyketide chain assembly by the rifF strain. Since a similar percentage decrease was seen in both genetic backgrounds, we conclude that the RifR thioesterase 2 is not involved in premature release of the carbon chain assembly intermediates. Similarly, fusion of the Saccharopolyspora erythraea DEBS3 thioesterase I domain to the C-terminus of the RifE PKS subunit did not result in a noticeable increase in the amount of the undecaketide intermediate formed nor in the amounts of the tetra- to decaketide shunt products. Hence, premature release of the carbon chain assembly intermediates is an unusual property of the Rif PKS itself.

There is currently intense interest in unravelling the modus operandi of type I modular polyketide synthases in order to lay the ground work for their use in the combinatorial biosynthesis of new bioactive molecules. Much of our knowledge is derived from studies on 6-deoxyerythronolide B (DEBS), the enzyme assembling the polyketide backbone of erythromycin. Work on the rifamycin polyketide synthase has revealed a number of features that differ from those seen with DEBS.

Malonate decarboxylases, which catalyze the conversion of malonate to acetate, can be classified into biotin-dependent and biotin-independent enzymes. In order to reveal the stereochemical course of the reactions catalyzed by the biotin-independent enzymes from Acinetobacter calcoaceticus and Pseudomonas fluorescens, a chiral substrate, malonate carrying (13)C in one carboxyl group and (3)H at one of the methylene positions, was prepared and used in the reactions catalyzed by these two enzymes. The decarboxylation of (R)-[1-(13)C(1), 2-(3)H]malonate in (2)H(2)O gave a pseudo-racemate of chiral acetate which was converted via acetyl-CoA into malate with malate synthase. From the relative proportions of the isotopomers of malate present, determined by (3)H NMR analysis, it was concluded that in the decarboxylation of malonate by these two biotin-independent enzymes COOH is replaced by H with retention of configuration. The same stereochemical outcome had been previously observed for the reaction catalyzed by the biotin-dependent malonate decarboxylase from Malonomonas rubra (J. Micklefield et al. J. Am. Chem. Soc. 117, 1153-1154, 1995)., (Copyright 1999 Academic Press.)

The biosynthesis of ansamycin antibiotics, including rifamycin B, involves the synthesis of an aromatic precursor, 3-amino-5-hydroxybenzoic acid (AHBA), which serves as starter for the assembly of the antibiotics' polyketide backbone. The terminal enzyme of AHBA formation, AHBA synthase, is a dimeric, pyridoxal 5'-phosphate (PLP) dependent enzyme with pronounced sequence homology to a number of PLP enzymes involved in the biosynthesis of antibiotic sugar moieties. The structure of AHBA synthase from Amycolatopsis mediterranei has been determined to 2.0 A resolution, with bound cofactor, PLP, and in a complex with PLP and an inhibitor (gabaculine). The overall fold of AHBA synthase is similar to that of the aspartate aminotransferase family of PLP-dependent enzymes, with a large domain containing a seven-stranded beta-sheet surrounded by alpha-helices and a smaller domain consisting of a four-stranded antiparallel beta-sheet and four alpha-helices. The uninhibited form of the enzyme shows the cofactor covalently linked to Lys188 in an internal aldimine linkage. On binding the inhibitor, gabaculine, the internal aldimine linkage is broken, and a covalent bond is observed between the cofactor and inhibitor. The active site is composed of residues from two subunits of AHBA synthase, indicating that AHBA synthase is active as a dimer.

The assembly of the polyketide backbone of rifamycin B on the type I rifamycin polyketide synthase (PKS), encoded by the rifA-rifE genes, is terminated by the product of the rifF gene, an amide synthase that releases the completed undecaketide as its macrocyclic lactam. Inactivation of rifF gives a rifamycin B nonproducing mutant that still accumulates a series of linear polyketides ranging from the tetra- to a decaketide, also detected in the wild type, demonstrating that the PKS operates in a processive manner. Disruptions of the rifD module 8 and rifE module 9 and module 10 genes also result in accumulation of such linear polyketides as a consequence of premature termination of polyketide assembly. Whereas the tetraketide carries an unmodified aromatic chromophore, the penta- through decaketides have undergone oxidative cyclization to the naphthoquinone, suggesting that this modification occurs during, not after, PKS assembly. The structure of one of the accumulated compounds together with (18)O experiments suggests that this oxidative cyclization produces an 8-hydroxy-7, 8-dihydronaphthoquinone structure that, after the stage of proansamycin X, is dehydrogenated to an 8-hydroxynaphthoquinone.

The putative biosynthetic gene cluster for the alpha-glucosidase inhibitor acarbose was identified in the producer Actinoplanes sp. 50/110 by cloning a DNA segment containing the conserved gene for dTDP-D-glucose 4,6-dehydratase, acbB. The two flanking genes were acbA (dTDP-D-glucose synthase) and acbC, encoding a protein with significant similarity to 3-dehydroquinate synthases (AroB proteins). The acbC gene was overexpressed heterologously in Streptomyces lividans 66, and the product was shown to be a C7-cyclitol synthase using sedo-heptulose 7-phosphate, but not ido-heptulose 7-phosphate, as its substrate. The cyclization product, 2-epi-5-epi-valiolone ((2S,3S,4S,5R)-5-(hydroxymethyl)cyclohexanon-2,3,4,5-tetrol), is a precursor of the valienamine moiety of acarbose. A possible five-step reaction mechanism is proposed for the cyclization reaction catalyzed by AcbC based on the recent analysis of the three-dimensional structure of a eukaryotic 3-dehydroquinate synthase domain (Carpenter, E. P., Hawkins, A. R., Frost, J. W., and Brown, K. A. (1998) Nature 394, 299-302).