Enzymatic synthesis of a bacterial polyketide from acetyl and malonyl coenzyme A.

Microorganisms and plants manufacture a large collection of medically and commercially useful natural products called polyketides by a process that resembles fatty acid biosynthesis. Genetically engineered microorganisms with modified polyketide synthase (PKS) genes can produce new metabolites that may have new or improved pharmacological activity. A potentially general method to prepare cell-free systems for studying bacterial type II PKS enzymes has been developed that facilitates the purification and reconstitution of their constituent proteins. Selective expression of different combinations of the Streptomyces glaucescens tetracenomycin (Tcm) tcmJKLMN genes in a tcmGHIJKLMNO null background has been used to show that the Tcm PKS consists of at least the TcmKLMN proteins. Addition of the TcmJ protein to the latter four enzymes resulted in a greater than fourfold increase of overall activity and thus represents the optimal Tcm PKS. Polyclonal antibodies raised against each of the TcmKLMN proteins strongly inhibit the Tcm PKS, as do known inhibitors targeted to the active site Cys and Ser residues of a fatty acid synthase. This system exhibits a strict starter unit specificity because neither propionyl, butyryl, or isobutyryl coenzyme A substitute for acetyl coenzyme A in assembly of the Tcm decaketide. Because the Tcm PKS activity is significantly diminished by removal of the TcmM acyl carrier protein and can be restored by addition of separately purified TcmM to two different types of TcmM-deficient PKS, it should be possible to use such preparations to assay for each of the constituents of the Tcm PKS.

Precursor-directed biosynthesis of erythromycin analogs by an engineered polyketide synthase.

A genetic block was introduced in the first condensation step of the polyketide biosynthetic pathway that leads to the formation of 6-deoxyerythronolide B (6-dEB), the macrocyclic precursor of erythromycin. Exogenous addition of designed synthetic molecules to small-scale cultures of this null mutant resulted in highly selective multimilligram production of unnatural polyketides, including aromatic and ring-expanded variants of 6-dEB. Unexpected incorporation patterns were observed, illustrating the catalytic versatility of modular polyketide synthases. Further processing of some of these scaffolds by postpolyketide enzymes of the erythromycin pathway resulted in the generation of novel antibacterials with in vitro potency comparable to that of their natural counterparts.

Modulation of polyketide synthase activity by accessory proteins during lovastatin biosynthesis.

Polyketides, the ubiquitous products of secondary metabolism in microorganisms, are made by a process resembling fatty acid biosynthesis that allows the suppression of reduction or dehydration reactions at specific biosynthetic steps, giving rise to a wide range of often medically useful products. The lovastatin biosynthesis cluster contains two type I polyketide synthase genes. Synthesis of the main nonaketide-derived skeleton was found to require the previously known iterative lovastatin nonaketide synthase (LNKS), plus at least one additional protein (LovC) that interacts with LNKS and is necessary for the correct processing of the growing polyketide chain and production of dihydromonacolin L. The noniterative lovastatin diketide synthase (LDKS) enzyme specifies formation of 2-methylbutyrate and interacts closely with an additional transesterase (LovD) responsible for assembling lovastatin from this polyketide and monacolin J.

Production of the antitumor drug epirubicin (4'-epidoxorubicin) and its precursor by a genetically engineered strain of Streptomyces peucetius.

A fermentation method that bypasses the low-yielding semisynthesis of epirubicin (4'-epidoxorubicin) and 4'-epidaunorubicin, important cancer chemotherapy drugs, has been developed for Streptomyces peucetius. This bacterium normally produces the anthracycline antibiotics, doxorubicin and daunorubicin; the 4'-epimeric anthracyclines are formed by introducing the heterologous Streptomyces avermitilis avrE or Saccharopolyspora eryBIV genes into an S. peucetius dnmV mutant blocked in the biosynthesis of daunosamine, the deoxysugar component of these antibiotics. Product yields were enhanced considerably by replacing the chromosomal copy of dnmV with avrE and by introducing further mutations that can increase daunorubicin and doxorubicin yields in the wild-type strain. This method demonstrates that valuable hybrid antibiotics can be made by combinatorial biosynthesis with bacterial deoxysugar biosynthesis genes.

Combinatorial biosynthesis for new drug discovery.

Combinatorial biosynthesis involves interchanging secondary metabolism genes between antibiotic-producing microorganisms to create unnatural gene combinations or hybrid genes if only part of a gene is exchanged. Novel metabolites can be made by both approaches, due to the effect of a new enzyme on a metabolic pathway or to the formation of proteins with new enzymatic properties. The method has been particularly successful with polyketide synthase (PKS) genes: derivatives of medically important macrolide antibiotics and unusual polycyclic aromatic compounds have been produced by novel combinations of the type I and type II PKS genes, respectively. Recent extensions of the approach to include deoxysugar biosynthesis genes have expanded the possibilities for making new microbial metabolites and discovering valuable drugs through the genetic engineering of bacteria.

Iterative type II polyketide synthases, cyclases and ketoreductases exhibit context-dependent behavior in the biosynthesis of linear and angular decapolyketides.

BACKGROUND: Iterative type II polyketide synthases (PKSs) produce polyketide chains of variable but defined length from a specific starter unit and a number of extender units. They also specify the initial regiospecific folding and cyclization pattern of nascent polyketides either through the action of a cyclase (CYC) subunit or through the combined action of site-specific ketoreductase (KR) and CYC subunits. Additional CYCs and other modifications may be necessary to produce linear aromatic polyketides. The principles of the assembly of the linear aromatic polyketides, several of which are medically important, are well understood, but it is not clear whether the assembly of the angular aromatic (angucyclic) polyketides follows the same rules. RESULTS: We performed an in vivo evaluation of the subunits of the PKS responsible for the production of the angucyclic polyketide jadomycin (jad), in comparison with their counterparts from the daunorubicin (dps) and tetracenomycin (tcm) PKSs which produce linear aromatic polyketides. No matter which minimal PKS was used to produce the initial polyketide chain, the JadD and DpsF CYCs produced the same two polyketides, in the same ratio; neither product was angularly fused. The set of jadABCED PKS plus putative jadl CYC genes behaved similarly. Furthermore, no angular polyketides were isolated when the entire set of jad PKS enzymes and Jadl or the jad minimal PKS, Jadl and the TcmN CYC were present. The DpsE KR was able to reduce decaketides but not octaketides; in contrast, the KRs from the jad PKS (JadE) or the actinorhodin PKS (ActIII) could reduce octaketide chains, giving three distinct products. CONCLUSIONS: It appears that the biosynthesis of angucyclic polyketides cannot be simply accomplished by expressing the known PKS subunits from artificial gene cassettes under the control of a non-native promoter. The characteristic structure of the angucycline ring system may arise from a kinked precursor during later cyclization reactions involving additional, but so far unknown, components of the extended decaketide PKS. Our results also suggest that some KRs have a minimal chain length requirement and that CYC enzymes may act aberrantly as first-ring aromatases that are unable to perform all of the sequential cyclization steps. Both of these characteristics may limit the widespread application of CYC or KR enzymes in the synthesis of novel polyketides.

A two-plasmid system for the glycosylation of polyketide antibiotics: bioconversion of epsilon-rhodomycinone to rhodomycin D.

BACKGROUND: The biological activity of many microbial products requires the presence of one or more deoxysugar molecules attached to agylcone. This is especially prevalent among polyketides and is an important reason that the antitumor anthracycline antibiotics are avid DNA-binding drugs. The ability to make different deoxyaminosugars and attach them to the same or different aglycones in vivo would facilitate the synthesis of new anthracyclines and the quest for antitumor drugs. This is feasible using the numerous bacterial genes for deoxysugar biosynthesis that are now available. RESULTS: Production of thymidine diphospho (TDP)-L-daunosamine (dnm), the aminodeoxysugar present in the anthracycline antitumor drugs daunorubicin (DNR) and doxorubicin (DXR), and its attachment to epsilon-rhodomycinone to generate rhodomycin D has been achieved by bioconversion with a strain of Streptomyces lividans that bears two plasmids. One contained the Streptomyces peucetius dnmJVUZTQS genes plus dnmW (previously named dpsH and considered to be a polyketide cyclase gene), dnrH, which is not required for the formation of rhodomycin D, and dnrI, a regulatory gene required for expression of the dnm and drr genes. The other plasmid had genes encoding glucose-1-phosphate thymidylyltransferase and TDP-glucose-4,6-dehydratase (dnmL and dnmM, respectively, or mtmDE, their homologs from Streptomyces agrillaceus) plus the drrAB DNR/DXR resistance genes. CONCLUSIONS: The high-yielding glycosylation of the aromatic polyketide epsilon-rhodomycinone using plasmid-borne deoxysugar biosynthesis genes proves that the minimal information for L-daunosamine biosynthesis and attachment in the heterologous host is encoded by the dnmLMJVUTS genes. This is a general approach to making both known and new glycosides of anthracyclines, several of which have medically important antitumor activity.

Engineered biosynthesis of geldanamycin analogs for Hsp90 inhibition.

Geldanamycin, a polyketide natural product, is of significant interest for development of new anticancer drugs that target the protein chaperone Hsp90. While the chemically reactive groups of geldanamycin have been exploited to make a number of synthetic analogs, including 17-allylamino-17-demethoxy geldanamycin (17-AAG), currently in clinical evaluation, the "inert" groups of the molecule remain unexplored for structure-activity relationships. We have used genetic engineering of the geldanamycin polyketide synthase (GdmPKS) gene cluster in Streptomyces hygroscopicus to modify geldanamycin at such positions. Substitutions of acyltransferase domains were made in six of the seven GdmPKS modules. Four of these led to production of 2-desmethyl, 6-desmethoxy, 8-desmethyl, and 14-desmethyl derivatives, including one analog with a four-fold enhanced affinity for Hsp90. The genetic tools developed for geldanamycin gene manipulation will be useful for engineering additional analogs that aid the development of this chemotherapeutic agent.

Insights about the biosynthesis of the avermectin deoxysugar L-oleandrose through heterologous expression of Streptomyces avermitilis deoxysugar genes in Streptomyces lividans.

BACKGROUND: The avermectins, produced by Streptomyces avermitilis, are potent anthelminthic agents with a polyketide-derived macrolide skeleton linked to a disaccharide composed of two alpha-linked L-oleandrose units. Eight contiguous genes, avrBCDEFGHI (also called aveBI-BVIII), are located within the avermectin-producing gene cluster and have previously been mapped to the biosynthesis and attachment of thymidinediphospho-oleandrose to the avermectin aglycone. This gene cassette provides a convenient way to study the biosynthesis of 2,6-dideoxysugars, namely that of L-oleandrose, and to explore ways to alter the biosynthesis and structures of the avermectins by combinatorial biosynthesis. RESULTS: A Streptomyces lividans strain harboring a single plasmid with the avrBCDEFGHI genes in which avrBEDC and avrIHGF were expressed under control of the actI and actIII promoters, respectively, correctly glycosylated exogenous avermectin A1a aglycone with identical oleandrose units to yield avermectin A1a. Modified versions of this minimal gene set produced novel mono- and disaccharide avermectins. The results provide further insight into the biosynthesis of L-oleandrose. CONCLUSIONS: The plasmid-based reconstruction of the avr deoxysugar genes for expression in a heterologous system combined with biotransformation has led to new information about the mechanism of 2,6-deoxysugar biosynthesis. The structures of the di-demethyldeoxysugar avermectins accumulated indicate that in the oleandrose pathway the stereochemistry at C-3 is ultimately determined by the 3-O-methyltransferase and not by the 3-ketoreductase or a possible 3,5-epimerase. The AvrF protein is therefore a 5-epimerase and not a 3,5-epimerase. The ability of the AvrB (mono-)glycosyltransferase to accommodate different deoxysugar intermediates is evident from the structures of the novel avermectins produced.

Biosynthesis of the ansamycin antibiotic rifamycin: deductions from the molecular analysis of the rif biosynthetic gene cluster of Amycolatopsis mediterranei S699.

BACKGROUND: The ansamycin class of antibiotics are produced by various Actinomycetes. Their carbon framework arises from the polyketide pathway via a polyketide synthase (PKS) that uses an unusual starter unit. Rifamycin (rif), produced by Amycolatopsis mediterranei, is the archetype ansamycin and it is medically important. Although its basic precursors (3-amino-5-hydroxy benzoic acid AHBA, and acetic and propionic acids) had been established, and several biosynthetic intermediates had been identified, very little was known about the origin of AHBA nor had the PKS and the various genes and enzymes that modify the initial intermediate been characterized. RESULTS: A set of 34 genes clustered around the rifK gene encoding AHBA synthase were defined by sequencing all but 5 kilobases (kb) of a 95 kb contiguous region of DNA from A. mediterranei. The involvement of some of the genes in the biosynthesis of rifamycin B was examined. At least five genes were shown to be essential for the synthesis of AHBA, five genes were determined to encode the modular type I PKS that uses AHBA as the starter unit, and 20 or more genes appear to govern modification of the polyketide-derived framework, and rifamycin resistance and export. Putative regulatory genes were also identified. Disruption of the PKS genes at the end of rifA abolished rifamycin B production and resulted in the formation of P8/1-OG, a known shunt product of rifamycin biosynthesis, whereas disruption of the orf6 and orf9 genes, which may encode deoxysugar biosynthesis enzymes, had no apparent effect. CONCLUSIONS: Rifamycin production in A. mediterranei is governed by a single gene cluster consisting of structural, resistance and export, and regulatory genes. The genes characterized here could be modified to produce novel forms of the rifamycins that may be effective against rifamycin-resistant microorganisms.

Cloning and heterologous expression of the genes encoding nonspecific electron transport components for a cytochrome P450 system of Saccharopolyspora erythraea involved in erythromycin production.

The forA gene encoding a protein that can function as a NADH:ferredoxin oxidoreductase (For) has been cloned from Saccharopolyspora erythraea, the erythromycin A (ErA) producer. In a previous study For protein, together with the FdxA ferredoxin from the same organism, was shown to be able to reconstitute the cytochrome P450 system responsible for the hydroxylation of 6-deoxyerythronolide B, an intermediate of ErA biosynthesis. Nucleotide sequence data suggest that the cloned forA gene codes for For, the putative pyruvate dehydrogenase component, dihydrolipoamide dehydrogenase, or its close homolog. Overexpression of forA appeared to be toxic to Escherichia coli.

Cloning and characterization of the Saccharopolyspora erythraea fdxA gene encoding ferredoxin.

The Saccharopolyspora erythraea gene (fdxA) corresponding to a previously purified ferredoxin [Shafiee and Hutchinson, J. Bacteriol., 170 (1988) 1548-1553] was cloned using an oligodeoxyribonucleotide probe based on the N-terminal sequence of the ferredoxin. The nucleotide sequence of a 1.3-kb segment encompassing fdxA indicates that the corresponding protein, SeFdI, is 105 amino acids long, and very similar to other 7Fe ferredoxins. A partial open reading frame closely linked to fdxA was also detected. Disruption of fdxA was attempted by replacing the wild-type allele with an in vitro mutated copy. The failure to construct an fdxA mutant strain suggests that fdxA lies in an essential region of the S. erythraea chromosome.

Regulation of expression of the valine (branched-chain amino acid) dehydrogenase-encoding gene from Streptomyces coelicolor.

Expression of the Streptomyces coelicolor (Sc) valine (branched-chain amino acid) dehydrogenase-encoding gene (vdh) is regulated by valine, glucose and NH+4 at the transcriptional level. The results of assays for the level of accumulated vdh mRNA in the Sc J802 strain by primer extension experiments and for the level of catechol dioxygenase (XylE) activity in Sc J802 (vdh::xylE) transformants show that transcription of the vdh gene is induced approx. 2.5-fold by valine, as compared to asparagine as the sole nitrogen source in the presence of glucose as carbon source. Valine induction is repressed by glucose, as compared to glycerol as the carbon source, and by NH+4. Glucose catabolite repression is relieved in Sc M480, a glucose kinase (glkA) deletion mutant. This suggests that glucose repression of vdh and carbohydrate metabolism are due to the same mechanism in Sc, which involves glucose kinase.

Characterization of the Streptomyces peucetius ATCC 29050 genes encoding doxorubicin polyketide synthase.

The dps genes of Streptomyces peucetius, encoding daunorubicin (DNR)-doxorubicin (DXR) polyketide synthase (PKS), are largely within an 8.7-kb region of DNA that has been characterized by Southern analysis, and gene sequencing, mutagenesis and expression experiments. This region contains nine ORFs, many of whose predicted products are homologous to known PKS enzymes. Surprisingly, the gene encoding the DXR PKS acyl carrier protein is not in this region, but is located about 10 kb distant from the position it usually occupies in other gene clusters encoding type-II PKS. An in-frame deletion in the dpsB gene, encoding a putative subunit of the DXR PKS, resulted in loss of production of DXR and the known intermediates of its biosynthetic pathway, confirming that this gene and, by implication, the adjacent dps genes are required for DXR biosynthesis. This was verified by expression of the dps genes in the heterologous host, Streptomyces lividans, which resulted in the production of aklanonic acid, an early intermediate of DXR biosynthesis.

Characterization of the enzymatic domains in the modular polyketide synthase involved in rifamycin B biosynthesis by Amycolatopsis mediterranei.

Five clustered polyketide synthase (PKS) genes, rifA-rifE, involved in rifamycin (Rf) biosynthesis in Amycolatopsis mediterranei S699 have been cloned and sequenced (August, P.R. et al., 1998. Chem. Biol. 5, 69-79). The five multifunctional polypeptides constitute a type I modular PKS that contains ten modules, each responsible for a specific round of polyketide chain elongation. Sequence comparisons of the Rf PKS proteins with other prokaryotic modular PKSs elucidated the regions that have an important role in enzyme activity and specificity. The beta-ketoacyl:acyl carrier protein synthase (KS) domains show the highest degree of similarity between themselves (86-90%) and to other PKSs (78-85%) among all the constituent domains. Both malonyl-coenzyme A (MCoA) and methylmalonyl-coenzyme A (mMCoA) are substrates for chain elongation steps carried out by the Rf PKS. Since acyltransferase (AT) domains of modular PKSs can distinguish between these two substrates, comparison of the sequence of all ten AT domains of the Rf PKS with those found in the erythromycin (Er) (Donadio, S. and Katz, L., 1992. Gene 111, 51-60) and rapamycin (Rp) (Haydock, S. et al., 1995. FEBS Lett. 374, 246-248) PKSs revealed that the AT domains in module 2 of RifA and module 9 of RifE are specific for MCoA, whereas the other eight modules specify mMCoA. Dehydration of the beta-hydroxyacylthioester intermediates should occur during the reactions catalysed by module 4 of RifB and modules 9 and 10 of RifE, yet only the active site region of module 4 conforms closely to the dehydratase (DH) motifs in the Er and Rp PKSs. The DH domains of modules 9 and 10 diverge significantly from the consensus sequence defined by the Er and Rp PKSs, except for the active site His residues. Deletions in the DH active sites of module 1 in RifA and module 5 in RifB and in the N- and C-terminal regions of module 8 of RifD should inactivate these domains, and module 2 of RifA lacks a DH domain, all of which are consistent with the proposed biosynthesis of Rf. In contrast, module 6 of RifB and module 7 of RifC appear to contain intact DH domains even though DH activity is not apparently required in these modules. Module 2 of RifA lacks a beta-ketoacyl:acyl carrier protein reductase (KR) domain and the one in module 3 has an apparently inactive NADPH binding motif, similar to one found in the Er PKS, while the other eight KR domains of the Rf PKS should be functional. These observations are consistent with biosynthetic predictions. All the acyl carrier protein (ACP) domains, while clearly functional, nevertheless have active site signature sequences distinctive from those of the Er and Rp PKSs. Module 2 of RifA has only the core domains (KS, AT and ACP). The starter unit ligase (SUL) and ACP domains present in the N-terminus of RifA direct the selection and loading of the starter unit, 3-amino-5-hydroxybenzoic acid (AHBA), onto the PKS. AHBA is made by the products of several other genes in the Rf cluster through a variant of the shikimate pathway (August, P.R. et al., inter alia). RifF, produced by the gene immediately downstream of rifE, is thought to catalyse the intramolecular cyclization of the PKS product, thereby forming the ansamacrolide precursor of Rf B. 1998 Elsevier Science B.V.

Cloning of the genes encoding thymidine diphosphoglucose 4,6-dehydratase and thymidine diphospho-4-keto-6-deoxyglucose 3,5-epimerase from the erythromycin-producing Saccharopolyspora erythraea.

Genes involved in deoxysugar metabolism, encoding thymidine diphospho (TDP)-glucose 4,6-dehydratase (gdh) and a putative TDP-4-keto-6-deoxyglucose 3,5-epimerase (kde), were cloned from the erythromycin (Er)-producing Saccharopolyspora erythraea by means of an oligodeoxynucleotide corresponding to a segment of the purified Gdh protein. Determination of the nucleotide sequence established that kde lies 3' to gdh. The function of gdh was confirmed by an enzymatic assay following expression of the gene in Escherichia coli. Southern analysis indicated that Sa. erythraea contains only one copy of gdh and kde. It was not possible to establish whether these genes are required for Er biosynthesis, but they appear to be essential for cellular metabolism, since resolution of a partial diploid containing a wt and a disrupted copy of gdh always maintained the wt gene. These loci do not lie within or near the known boundaries of the cluster of Er-production and -resistance genes, nor do they appear to be flanked by other deoxysugar biosynthesis genes.

Identification of Streptomyces olivaceus Tü 2353 genes involved in the production of the polyketide elloramycin.

The genes for the production of elloramycin (ELM) from Streptomyces olivaceus (So) Tü2353 were cloned using a polyketide synthase gene probe from the tetracenomycin pathway. A cosmid clone (16F4) isolated from a gene library of So Tü2353 conferred tetracenomycin C and ELM resistance to S. lividans TK64 and complemented a mutation in So Tü2353R. Introduction of cosmid 16F4 into S. lividans TK64 resulted in the production of 8-demethyl-tetracenomycin C, an intermediate of ELM biosynthesis.

Combinatorial biosynthesis in microorganisms as a route to new antimicrobial, antitumor and neuroregenerative drugs.

Combinatorial biosynthesis utilizes the genes of biosynthetic pathways that produce microbial products to create novel chemical structures. The engineering of mondular polyketide synthase (PKS) genes has been the major focus of this effort and has led to the production of analogs of macrolide antibiotics like the erythromycins and their derived ketolides, and of the immunosuppressive macrolide FK-520 (Fujisawa Pharmaceutical Co Ltd). Approaches to making analogs of the promising antitumor compounds known as epothilones are also being explored. Lead compounds for further study have resulted and routes to analogs of other pharmacologically important compounds have been established. To facilitate this work, many new tools for manipulating and studying the multifunctional PKSs have been developed including the development of Escherichia coli as a PKS expression last. These developments have resulted in faster ways of engineering PKS to produce new compounds for the development of chemotherapeutic agents from natural products.

Triple hydroxylation of tetracenomycin A2 to tetracenomycin C involving two molecules of O(2) and one molecule of H(2)O.

The TcmG or ElmG oxygenase-catalyzed triple hydroxylation of tetracenomycin (Tcm) A2 to Tcm C proceeds via a novel monooxygenase-dioxygenase mechanism, deriving the 4- and 12a-OH groups of Tcm C from two molecules of O(2) and the 4a-OH group of Tcm C from a molecule of H(2)O. These results suggest a mechanistic analogy among TcmG, ElmG, and the bacterial and fungal hydroquinone epoxidizing dioxygenases, as well as the mammalian vitamin K-dependent gamma-glutamyl carboxylase.

The genetic basis of precursor supply for the biosynthesis of macrolide and polyether antibiotics.

Macrolide and polyether biosynthesis in actinomycetes is regulated at the level of precursor supply by effects of nutrients on the sources of the low-molecular-weight fatty acids used to build the carbon framework of these antibiotics. Ammonium ion appears to suppress the first enzymes of valine and threonine catabolism and also inhibits their activity. Disruption of the valine dehydrogenase (vdh) gene of Streptomyces coelicolor destroys its ability to grow on branched-chain amino acids as the sole nitrogen source in a minimal medium but has no effect on the biosynthesis of the acetate-derived antibiotic, actinorhodin. Expression of the vdh gene is repressed by > 25 mM ammonium ion or glucose but not by valine, glycerol, or maltose. Vdh enzyme activity is stimulated by valine induction. These results suggest that the inhibition of valine catabolism by ammonium and/or glucose could explain why macrolide production is inhibited by ammonium ion.