Characterization of Streptomyces argillaceus genes encoding a polyketide synthase involved in the biosynthesis of the antitumor mithramycin

Mithramycin (Mtm) is an aromatic polyketide which shows antibacterial and antitumor activity. From a chromosomal cosmid library of Streptomyces argillaceus, a Mtm producer, a clone (cosAR7) was isolated by homology to the actI/III region of S. coelicolor and the strDEM genes of S. griseus. From this clone, a 5.3-kb DNA region was sequenced and found to encode six open reading frames (designated as mtmQXPKSTI), five of them transcribed in the same direction. The deduced products of five of these genes resembled components of type-II polyketide synthases. The mtm genes would code for an aromatase (mtmQ), a polypeptide of unknown function (mtmX), a β-ketoacylsynthase (mtmP) and a related ‘chain length factor’ (mtmK), an acyl carrier protein (mtmS) and a β-ketoreductase (mtmT1). The involvement of this gene cluster in Mtm biosynthesis was demonstrated by the Mtm non-producing phenotype of mutants generated in two independent insertional inactivation experiments.     

Characterization of two glycosyltransferases involved in early glycosylation steps during biosynthesis of the antitumor polyketide mithramycin by Streptomyces argillaceus

A 2580-bp region of the chromosome of Streptomyces argillaceus, the producer of the antitumor polyketide mithramycin, was sequenced. Analysis of the nucleotide sequence revealed the presence of two genes (mtmGIII and mtmGIVv) encoding proteins that showed a high degree of similarity to glycosyltransferases involved in the biosynthesis of various antibiotics and antitumor drugs. Independent insertional inactivation of both genes produced mutants that did not synthesize mithramycin but accumulated several mithramycin intermediates. Both mutants accumulated premithramycinone, a non-glycosylated intermediate in mithramycin biosynthesis. The mutant affected in the mtmGIII gene also accumulated premithramycin A1, which contains premithramycinone as the aglycon unit and a D-olivose attached at C-12a-O. These experiments demonstrate that the glycosyltransferases MtmGIV and MtmGIII catalyze the first two glycosylation steps in mithramycin biosynthesis. A model is proposed for the glycosylation steps in mithramycin biosynthesis.     

Characterization of two polyketide methyltransferases involved in the biosynthesis of the antitumor drug mithramycin by Streptomyces argillaceus

A DNA chromosomal region of Streptomyces argillaceus ATCC 12596, the producer organism of the antitumor polyketide drug mithramycin, was cloned. Sequence analysis of this DNA region, located between four mithramycin glycosyltransferase genes, showed the presence of two genes (mtmMI and mtmMII) whose deduced products resembled S-adenosylmethionine-dependent methyltransferases. By independent insertional inactivation of both genes nonproducing mutants were generated that accumulated different mithramycin biosynthetic intermediates. The M3DeltaMI mutant (mtmMI-minus mutant) accumulated 4-demethylpremithramycinone (4-DPMC) which lacks the methyl groups at carbons 4 and 9. The M3DeltaM2 (mtmMII-minus mutant) accumulated 9-demethylpremithramycin A3 (9-DPMA3), premithramycin A1 (PMA1), and 7-demethylmithramycin, all of them containing the O-methyl group at C-4 and C-1', respectively, but lacking the methyl group at the aromatic position. Both genes were expressed in Streptomyces lividans TK21 under the control of the erythromycin resistance promoter (ermEp) of Saccharopolyspora erythraea. Cell-free extracts of these clones were precipitated with ammonium sulfate (90% saturation) and assayed for methylation activity using different mithramycin intermediates as substrates. Extracts of strains MJM1 (expressing the mtmMI gene) and MJM2 (expressing the mtmMII gene) catalyzed efficient transfer of tritium from [(3)H]S-adenosylmethionine into 4-DPMC and 9-DPMA3, respectively, being unable to methylate other intermediates at a detectable level. These results demonstrate that the mtmMI and mtmMII genes code for two S-adenosylmethionine-dependent methyltransferases responsible for the 4-O-methylation and 9-C-methylation steps of the biosynthetic precursors 4-DPMC and 9-DPMA3, respectively, of the antitumor drug mithramycin. A pathway is proposed for the last steps in the biosynthesis of mithramycin involving these methylation events.     

Deoxysugar transfer during chromomycin A3 biosynthesis in Streptomyces griseus subsp. griseus: new derivatives with antitumor activity

Chromomycin A3 is an antitumor drug produced by Streptomyces griseus subsp. griseus. It consists of a tricyclic aglycone with two aliphatic side chains and two O-glycosidically linked saccharide chains, a disaccharide of 4-O-acetyl-D-oliose (sugar A) and 4-O-methyl-D-oliose (sugar B), and a trisaccharide of D-olivose (sugar C), D-olivose (sugar D), and 4-O-acetyl-L-chromose B (sugar E). The chromomycin gene cluster contains four glycosyltransferase genes (cmmGI, cmmGII, cmmGIII, and cmmGIV), which were independently inactivated through gene replacement, generating mutants C60GI, C10GII, C10GIII, and C10GIV. Mutants C10GIV and C10GIII produced the known compounds premithramycinone and premithramycin A1, respectively, indicating the involvement of CmmGIV and CmmGIII in the sequential transfer of sugars C and D and possibly also of sugar E of the trisaccharide chain, to the 12a position of the tetracyclic intermediate premithramycinone. Mutant C10GII produced two new tetracyclic compounds lacking the disaccharide chain at the 8 position, named prechromomycin A3 and prechromomycin A2. All three compounds accumulated by mutant C60GI were tricyclic and lacked sugar B of the disaccharide chain, and they were named prechromomycin A4, 4A-O-deacetyl-3A-O-acetyl-prechromomycin A4, and 3A-O-acetyl-prechromomycin A4. CmmGII and CmmGI are therefore responsible for the formation of the disaccharide chain by incorporating, in a sequential manner, two D-oliosyl residues to the 8 position of the biosynthetic intermediate prechromomycin A3. A biosynthetic pathway is proposed for the glycosylation events in chromomycin A3 biosynthesis.     

Minimal Streptomyces sp. strain C5 daunorubicin polyketide biosynthesis genes required for aklanonic acid biosynthesis.

The structure of the Streptomyces sp. strain C5 daunorubicin type II polyketide synthase (PKS) gene region is different from that of other known type II PKS gene clusters. Directly downstream of the genes encoding ketoacylsynthase alpha and beta (KS alpha, KS beta) are two genes (dpsC, dpsD) encoding proteins of unproven function, both absent from other type II PKS gene clusters. Also in contrast to other type II PKS clusters, the gene encoding the acyl carrier protein (ACP), dpsG, is located about 6.8 kbp upstream of the genes encoding the daunorubicin KS alpha and KS beta. In this work, we demonstrate that the minimal genes required to produce aklanonic acid in heterologous hosts are dpsG (ACP), dauI (regulatory activator), dpsA (KS alpha), dpsB (KS beta), dpsF (aromatase), dpsE (polyketide reductase), and dauG (putative deoxyaklanonic acid oxygenase). The two unusual open reading frames, dpsC (KASIII homolog lacking a known active site) and dpsD (acyltransferase homolog), are not required to synthesize aklanonic acid. Additionally, replacement of dpsD or dpsCD in Streptomyces sp. strain C5 with a neomycin resistance gene (aphI) results in mutant strains that still produced anthracyclines.     

Analysis of two chromosomal regions adjacent to genes for a type II polyketide synthase involved in the biosynthesis of the antitumor polyketide mithramycin in Streptomyces argillaceus

Mithramycin is an aromatic antitumour polyketide synthesized by Streptomyces argillaceus. Two chromosomal regions located upstream and downstream of the locus for the mithramycin type II polyketide synthase were cloned and sequenced. Analysis of the sequence revealed the presence of eight genes encoding three oxygenases (mtmOI, mtmOII and mtmOIII), three reductases (mtmTI, mtmTII and mtmTIII), a cyclase (mtmY) and an acyl CoA ligase (mtmL). The three oxygenase genes were each inactivated by gene replacement. Inactivation of one of them (mtmOII) generated a non-producing mutant, while inactivation of the other two (mtmOI and mtmOIII) did not affect the biosynthesis of mithramycin. The mtmOII gene may code for an oxygenase responsible for the introduction of oxygen atoms at early steps in the biosynthesis of mithramycin leading to 4-demethylpremithramycinone. One of the reductases may be responsible for reductive cleavage of an intermediate from an enzyme and another for the reduction of a keto group in the side-chain of the mithramycin aglycon moiety. A hypothetical biosynthetic pathway showing in particular the involvement of oxygenase MtmOII and of various other gene products in mithramycin biosynthesis is proposed. Received: 13 August 1998 / Accepted: 30 October 1998     

Cloning and insertional inactivation of Streptomyces argillaceus genes involved in the earliest steps of biosynthesis of the sugar moieties of the antitumor polyketide mithramycin.

Two genes (mtmD and mtmE) were cloned and sequenced from the mithramycin producer Streptomyces argillaceus. Comparison with proteins in databases and enzymatic assays after expression in Escherichia coli showed that they encode a glucose-1-phosphate:TTP thymidylyl transferase and a TDP-D-glucose 4,6-dehydratase, respectively. The mtmD gene was inactivated by gene replacement, generating a nonproducing mutant that accumulates a tetracyclic compound designated premithramycinone. The identification of premithramycinone reveals new aspects of the mithramycin biosynthetic pathway and suggests that at least some glycosylations occur before breakage of the fourth ring.     

Characterization of two glycosyltransferases involved in early glycosylation steps during biosynthesis of the antitumor polyketide mithramycin by <Emphasis Type="Italic">Streptomyces argillaceus</Emphasis>

A 2580-bp region of the chromosome of Streptomyces argillaceus, the producer of the antitumor polyketide mithramycin, was sequenced. Analysis of the nucleotide sequence revealed the presence of two genes (mtmGIII and mtmGIV?) encoding proteins that showed a high degree of similarity to glycosyltransferases involved in the biosynthesis of various antibiotics and antitumor drugs. Independent insertional inactivation of both genes produced mutants that did not synthesize mithramycin but accumulated several mithramycin intermediates. Both mutants accumulated premithramycinone, a non-glycosylated intermediate in mithramycin biosynthesis. The mutant affected in the mtmGIII gene also accumulated premithramycin A1, which contains premithramycinone as the aglycon unit and a D-olivose attached at C-12a-O. These experiments demonstrate that the glycosyltransferases MtmGIV and MtmGIII catalyze the first two glycosylation steps in mithramycin biosynthesis. A model is proposed for the glycosylation steps in mithramycin biosynthesis.     

Combinatorial polyketide biosynthesis by de novo design and rearrangement of modular polyketide synthase genes

Type I polyketide synthase (PKS) genes consist of modules approximately 3-6 kb long, which encode the structures of 2-carbon units in polyketide products. Alteration or replacement of individual PKS modules can lead to the biosynthesis of 'unnatural' natural products but existing techniques for this are time consuming. Here we describe a generic approach to the design of synthetic PKS genes where facile cassette assembly and interchange of modules and domains are facilitated by a repeated set of flanking restriction sites. To test the feasibility of this approach, we synthesized 14 modules from eight PKS clusters and associated them in 154 bimodular combinations spanning over 1.5-million bp of novel PKS gene sequences. Nearly half the combinations successfully mediated the biosynthesis of a polyketide in Escherichia coli, and all individual modules participated in productive bimodular combinations. This work provides a truly combinatorial approach for the production of polyketides.     

Characterization of two polyketide synthase genes involved in zearalenone biosynthesis in Gibberella zeae

Zearalenone, a mycotoxin produced by several Fusarium spp., is most commonly found as a contaminant in stored grain and has chronic estrogenic effects on mammals. Zearalenone is a polyketide derived from the sequential condensation of multiple acetate units by a polyketide synthase (PKS), but the genetics of its biosynthesis are not understood. We cloned two genes, designated ZEA1 and ZEA2, which encode polyketide synthases that participate in the biosynthesis of zearalenone by Gibberella zeae (anamorph Fusarium graminearum). Disruption of either gene resulted in the loss of zearalenone production under inducing conditions. ZEA1 and ZEA2 are transcribed divergently from a common promoter region. Quantitative PCR analysis of both PKS genes and six flanking genes supports the view that the two polyketide synthases make up the core biosynthetic unit for zearalenone biosynthesis. An appreciation of the genetics of zearalenone biosynthesis is needed to understand how zearalenone is synthesized under field conditions that result in the contamination of grain.     

Complex organization of the Streptomyces avermitilis genes encoding the avermectin polyketide synthase.

The avermectin (Av) polyketide synthase (PKS) and erythromycin (Er) PKS are encoded by modular repeats of DNA, but the genetic organization of the modules encoding Av PKS is more complex than Er PKS. Sequencing of several related DNA fragments from Streptomyces avermitilis that are part of the Av biosynthetic gene cluster, revealed that they encode parts of large multifunctional PKS proteins. The Av PKS proteins show strong similarity to each other, as well as similarity to Er PKS proteins [Donadio et al., Science 252 (1991) 675-679] and fatty acid synthases. Partial DNA sequencing of the 65-kb region containing all the related sequence elements in the avr genes provides evidence for twelve modular repeats encoding FAS-like domains. The genes encoding the Av PKS are organized as two sets of six modular repeats which are convergently transcribed.     

Characterization of tailoring genes involved in the modification of geldanamycin polyketide in Streptomyces hygroscopicus JCM4427

Geldanamycin and its analogs are important anticancer agents that inhibit the newly targeted, heat-shock protein (Hsp) 90, which is a chaperone protein in eukaryotic cells. To resolve which geldanamycin biosynthetic genes are responsible for particular post-polyketide synthase (PKS) processing steps and in which order the reactions occur, we individually inactivated candidate genes in Streptomyces hygroscopicus subsp. duamyceticus JCM4427, and isolated and elucidated the structures of intermediates from each mutant. The results indicated that gel7 governs at least one of the benzoquinone ring oxidation steps. In addition, gel16 was found to be involved in double-bond formation between C-4 and C-5 of 4,5-dihydrogeldanamycin, which confirmed our previous findings that this double bond reduced during the post-PKS modification of the polyketide assembly. In addition, pro-geldanamycin, which does not possess a double bond at C-4/5, was purified from the gel7 and 8 double-gene-inactivated mutant.     

Generation of hybrid elloramycin analogs by combinatorial biosynthesis using genes from anthracycline-type and macrolide biosynthetic pathways

Elloramycin and oleandomycin are two polyketide compounds produced by Streptomyces olivaceus Tü2353 and Streptomyces antibioticus ATCC11891, respectively. Elloramycin is an anthracycline-like antitumor drug and oleandomycin a macrolide antibiotic. Expression in S. albus of a cosmid (cos16F4) containing part of the elloramycin biosynthetic gene cluster produced the elloramycin non-glycosylated intermediate 8-demethyl-tetracenomycin C. Several plasmid constructs harboring different gene combinations of L-oleandrose (neutral 2,6-dideoxyhexose attached to the macrolide antibiotic oleandomycin) biosynthetic genes of S. antibioticus that direct the biosynthesis of L-olivose, L-oleandrose and L-rhamnose were coexpressed with cos16F4 in S. albus. Three new hybrid elloramycin analogs were produced by these recombinant strains through combinatorial biosynthesis, containing elloramycinone or 12a-demethyl-elloramycinone (= 8-demethyl-tetracenomycin C) as aglycone moiety encoded by S. olivaceus genes and different sugar moieties, coded by the S. antibioticus genes. Among them is L-olivose, which is here described for the first time as a sugar moiety of a natural product.     

Analysis of the nucleotide sequence of the Streptomyces glaucescens tcmI genes provides key information about the enzymology of polyketide antibiotic biosynthesis.

Key information about the biosynthesis of polyketide metabolites has been uncovered by sequence analysis of the tetracenomycin C polyketide synthase genes (tcml) from Streptomyces glaucescens GLA.0. The sequence data revealed the presence of three complete open reading frames (ORFs). ORF1 and ORF2 appear to be translationally coupled and would encode proteins containing 426 and 405 amino acids, respectively. The two deduced proteins are homologous to known beta-ketoacyl synthases. ORF3 begins 70 nucleotides after the stop codon of ORF2 and would code for an 83 amino acid protein with a strong resemblance to known bacterial, animal and plant acyl-carrier proteins (ACP). The presence of an ACP gene within the tcm gene cluster suggests that different ACPs are used in fatty acid and polyketide biosynthesis in Streptomyces. We conclude from these data and earlier information that polyketide biosynthesis in S. glaucescens, and most likely in other bacteria, involves a multienzyme complex consisting of at least five types of enzymes: acylCoA transferases that load the acyl and 2-carboxyacyl precursors onto the ACP; a beta-ketoacyl synthase that, along with the acylated ACP, forms the poly-beta-ketoacyl intermediates; a poly-beta-ketone cyclase that forms carbocyclic structures from the latter intermediates; a beta-ketoacyl oxidoreductase that forms beta-hydroxyacyl intermediates or reduces ketone groups in fully formed polyketides; and a thioesterase that releases the assembled polyketide from the enzyme.     

Identification and characterization of the niddamycin polyketide synthase genes from Streptomyces caelestis.

The genes encoding the polyketide synthase (PKS) portion of the niddamycin biosynthetic pathway were isolated from a library of Streptomyces caelestis NRRL-2821 chromosomal DNA. Analysis of 40 kb of DNA revealed the presence of five large open reading frames (ORFs) encoding the seven modular sets of enzymatic activities required for the synthesis of a 16-membered lactone ring. The enzymatic motifs identified within each module were consistent with those predicted from the structure of niddamycin. Disruption of the second ORF of the PKS coding region eliminated niddamycin production, demonstrating that the cloned genes are involved in the biosynthesis of this compound.     

The O-methyltransferase SrsB catalyzes the decarboxylative methylation of alkylresorcylic acid during phenolic lipid biosynthesis by Streptomyces griseus

Streptomyces griseus contains the srs operon, which is required for phenolic lipid biosynthesis. The operon consists of srsA, srsB, and srsC, which encode a type III polyketide synthase, an O-methyltransferase, and a flavoprotein hydroxylase, respectively. We previously reported that the recombinant SrsA protein synthesized 3-(13'-methyltetradecyl)-4-methylresorcinol, using iso-C(16) fatty acyl-coenzyme A (CoA) as a starter substrate and malonyl-CoA and methylmalonyl-CoA as extender substrates. An in vitro SrsA reaction using [(13)C(3)]malonyl-CoA confirmed that the order of extender substrate condensation was methylmalonyl-CoA, followed by two extensions with malonyl-CoA. Furthermore, SrsA was revealed to produce an alkylresorcylic acid as its direct product rather than an alkylresorcinol. The functional SrsB protein was produced in the membrane fraction in Streptomyces lividans and used for the in vitro SrsB reaction. When the SrsA reaction was coupled, SrsB produced alkylresorcinol methyl ether in the presence of S-adenosyl-l-methionine (SAM). SrsB was incapable of catalyzing the O-methylation of alkylresorcinol, indicating that alkylresorcylic acid was the substrate of SrsB and that SrsB catalyzed the conversion of alkylresorcylic acid to alkylresorcinol methyl ether, namely, by both the O-methylation of the hydroxyl group (C-6) and the decarboxylation of the neighboring carboxyl group (C-1). O-methylated alkylresorcylic acid was not detected in the in vitro SrsAB reaction, although it was presumably stable, indicating that O-methylation did not precede decarboxylation. We therefore postulated that O-methylation was coupled with decarboxylation and proposed that SrsB catalyzed the feasible SAM-dependent decarboxylative methylation of alkylresorcylic acid. To the best of our knowledge, this is the first report of a methyltransferase that catalyzes decarboxylative methylation.     

Repeated polyketide synthase modules involved in the biosynthesis of a heptaene macrolide by Streptomyces sp. FR-008

Genes for biosynthesis of a Streptomyces sp. FR-008 heptaene macrolide antibiotic with antifungal and mosquito larvicidal activity were cloned in Escherichia coli using heterologous DNA probes. The cloned genes were implicated in heptaene biosynthiesis by gene replacement. The FR-008 antibiotic contains a 38-membered, poiyketide-derived macrolide ring. Southern hybridization using probes encoding domains of the type i modular erythromycin polyketide synthase (PKS) showed that the Streptomyces sp. FR-008 PKS gene cluster contains repeated sequences spanning c. 105 kb of contiguous DNA; assuming c. 5 kb for each PKS module, this is in striking agreement with the expectation for the 21-step condensation process required for synthesis of the FR-008 carbon chain. The methods developed for transformation and gene replacement in Streptomyces sp. FR-008 make it possible to genetically manipulate polyene macrolide production, and may later lead to the biosynthesis of novel polyene macrolides.     

Characterization of methyltransferase and hydroxylase genes involved in the biosynthesis of the immunosuppressants FK506 and FK520.

FK506 and FK520 are 23-membered macrocyclic polyketides with potent immunosuppressive and antifungal activities. The gene encoding 31-O-demethyl-FK506 methyltransferase, fkbM, was isolated from Streptomyces sp. strains MA6858 and MA6548, two FK506 producers, and Streptomyces hygroscopicus subsp. ascomyceticus, an FK520 producer. The nucleotide sequence of the fkbM gene revealed an open reading frame encoding a polypeptide of 260 amino acids. Disruption of fkbM in Streptomyces sp. strain MA6548 yielded a mutant that produced 31-O-demethyl-FK506, confirming the involvement of the isolated genes in the biosynthesis of FK506 and FK520. Heterologous expression of fkbM in Streptomyces lividans established that fkbM encodes an O-methyltransferase catalyzing the methylation of the C-31 hydroxyl group of 31-O-demethyl-FK506 and FK520. A second open reading frame, fkbD, was found upstream of fkbM in all three aforementioned species and was predicted to encode a protein of 388 residues that showed a strong resemblance to cytochrome P-450 hydroxylases. Disruption of fkbD had a polar effect on the synthesis of the downstream fkbM gene product and resulted in the formation of 9-deoxo-31-O-demethyl-FK506. This established the product of fkbD as the cytochrome P-450 9-deoxo-FK506 hydroxylase, which is responsible for hydroxylation at position C-9 of the FK506 and FK520 macrolactone ring.     

Characterization of the groEL-like genes in Streptomyces albus.

Three GroEL-like heat shock proteins (HSP56, HSP58, and HSP18) have been observed in Streptomyces albus (G. Guglielmi, P. Mazodier, C. J. Thompson, and J. Davies, J. Bacteriol. 173:7374-7381, 1991). Here we report the cloning and complete nucleotide sequence of groEL1, which encodes HSP18 and HSP58, and groEL2, which encodes HSP56. Both nucleotide sequences predicted proteins of 56,680 Da that were 70% identical. The 5' nucleotide sequence of groEL1 coded for a protein corresponding to HSP18 that may be a processed gene product. At least two groEL-like genes were present in all 12 Streptomyces species tested; they were not closely linked in the genome. groEL1, but not groEL2, was adjacent to a groES-like gene.