Establishing a new methodology for genome mining and biosynthesis of polyketides and peptides through yeast molecular genetics

Fungal genome sequencing has revealed many genes coding for biosynthetic enzymes, including polyketide synthases and nonribosomal peptide synthetases. However, characterizing these enzymes and identifying the compounds they synthesize remains a challenge, whether the genes are expressed in their original hosts or in more tractable heterologous hosts, such as yeast. Here, we developed a streamlined method for isolating biosynthetic genes from fungal sources and producing bioactive molecules in an engineered Saccharomyces cerevisiae host strain. We used overlap extension PCR and yeast homologous recombination to clone desired fungal polyketide synthase or a nonribosomal peptide synthetase genes (5-20 kb) into a yeast expression vector quickly and efficiently. This approach was used successfully to clone five polyketide synthases and one nonribosomal peptide synthetase, from various fungal species. Subsequent detailed chemical characterizations of the resulting natural products identified six polyketide and two nonribosomal peptide products, one of which was a new compound. Our system should facilitate investigating uncharacterized fungal biosynthetic genes, identifying novel natural products, and rationally engineering biosynthetic pathways for the production of enzyme analogues possessing modified bioactivity.     

Identification of the Tirandamycin Biosynthetic Gene Cluster from Streptomyces sp. 307‐9

The structurally intriguing bicyclic ketal moiety of tirandamycin is common to several acyl‐tetramic acid antibiotics, and is a key determinant of biological activity. We have identified the tirandamycin biosynthetic gene cluster from the environmental marine isolate Streptomyces sp. 307–9, thus providing the first genetic insight into the biosynthesis of this natural product scaffold. Sequence analysis revealed a hybrid polyketide synthase–nonribosomal peptide synthetase gene cluster with a colinear domain organization, which is entirely consistent with the core structure of the tirandamycins. We also identified genes within the cluster that encode candidate tailoring enzymes for elaboration and modification of the bicyclic ketal system. Disruption of tamI, which encodes a presumed cytochrome P450, led to a mutant strain deficient in production of late stage tirandamycins that instead accumulated tirandamycin C, an intermediate devoid of any post assembly‐line oxidative modifications.     

Activity of α‐Aminoadipate Reductase Depends on the N‐Terminally Extending Domain

L ‐α‐Aminoadipic acid reductases catalyze the ATP‐ and NADPH‐dependent reduction of L ‐α‐aminoadipic acid to the corresponding 6‐semialdehyde during fungal L ‐lysine biosynthesis. These reductases resemble peptide synthetases with regard to their multidomain composition but feature a unique domain of elusive function—now referred to as an adenylation activating (ADA) domain—that extends the reductase N‐terminally. Truncated enzymes based on NPS3, the L ‐α‐aminoadipic acid reductase of the basidiomycete Ceriporiopsis subvermispora, lacking the ADA domain either partially or entirely were tested for activity in vitro, together with an ADA‐adenylation didomain and the ADA domainless adenylation domain. We provide evidence that the ADA domain is required for substrate adenylation: that is, the initial step of the catalytic turnover. Our biochemical data are supported by in silico modeling that identified the ADA domain as a partial peptide synthetase condensation domain.     

Insights into the Role of Ketoreductases in the Biosynthesis of Partially Reduced Bacterial Aromatic Polyketides*

Partially reduced aromatic polyketides are bioactive secondary metabolites or intermediates in the biosynthesis of deoxygenated aromatics. For the antibiotic GTRI-02 (mensalone) in different Streptomyces spp., biosynthesis involving the reduction of a fully aromatized acetyltrihydroxynaphthalene by a naphthol reductase has been proposed and shown in vitro with a fungal enzyme. However, more recently, GTRI-02 has been identified as a product of the ActIII biosynthetic gene cluster from Streptomyces coelicolor A3(2), for which the reduction of a linear polyketide precursor by ActIII ketoreductase, prior to cyclization and aromatization, has been suggested. We have examined three different ketoreductases from bacterial producer strains of GTRI-02 for their ability to reduce mono-, bi-, and tricyclic aromatic substrates. The enzymes reduced 1- and 2-tetralone but not other aromatic substrates. This strongly suggests a reduction of a cyclized but not yet aromatic polyketide intermediate in the biosynthesis of GTRI-02. Implications of the results for the biosynthesis of other secondary polyketidic metabolites are discussed.     

Evolutionary Imprint of Catalytic Domains in Fungal PKS–NRPS Hybrids

Fungal hybrid enzymes consisting of a polyketide synthase (PKS) and a nonribosomal peptide synthetase (NRPS) module are involved in the biosynthesis of a vast array of ecologically and medicinally relevant natural products. Whereas a dozen gene clusters could be assigned to the requisite PKS–NRPS pathways, the programming of the multifunctional enzymes is still enigmatic. Through engineering and heterologously expressing a chimera of PKS (lovastatin synthase, LovB) and NRPS (cytochalasin synthase, CheA) in Aspergillus terreus, we noted the potential incompatibility of a fungal highly reducing PKS (hrPKS) with the NRPS component of fungal PKS–NRPS hybrids. To rationalize the unexpected outcome of the gene fusion experiments, we conducted extensive bioinformatic analyses of fungal PKS–NRPS hybrids and LovB‐type PKS. From motif studies and the function of the engineered chimeras, a noncanonical function of C‐terminal condensation (C) domains in truncated PKS–NRPS homologues was inferred. More importantly, sequence alignments and phylogenetic trees revealed an evolutionary imprint of the PKS–NRPS domains, which reflect the evolutionary history of the entire megasynthase. Furthermore, a detailed investigation of C and adenylation (A) domains provides support for a scenario in which not only the A domain but also the C domain participates in amino acid selection. These findings shed new light on the complex code of this emerging class of multifunctional enzymes and will greatly facilitate future combinatorial biosynthesis and pathway engineering approaches towards natural product analogues.     

Identification and Characterization of the Cuevaene A Biosynthetic Gene Cluster in Streptomyces sp. LZ35

Genome sequence analysis of Streptomyces sp. LZ35 has revealed a large number of secondary metabolite pathways, including one encoded in an orphan type I polyketide synthase gene cluster that contains a putative chorismatase/3‐hydroxybenzoate synthase gene. Mutagenesis and comparative metabolic profiling led to the identification of cuevaene A as the metabolic product of the gene cluster, thus making it the first 3‐HBA containing polyketide biosynthetic gene cluster described to date. Cuv10 was proven to be responsible for the conversion of chorismate into 3‐HBA; Cuv18 is speculated to be responsible for the 6‐hydroxylation of 3‐HBA during polyketide chain elongation. Additionally, several pathway‐specific regulatory factors that affect the production of cuevaene A were identified. Our results indicate that targeted inactivation of a gene followed by comparative metabolic profiling is a useful approach to identify and characterize cryptic biosynthetic gene clusters.     

Synthesis of Dolichols in Candida albicans Is Co-Regulated with Elongation of Fatty Acids

Protein glycosylation requires dolichyl phosphate as a carbohydrate carrier. Dolichols are α-saturated polyprenols, and their saturation in S. cerevisiae is catalyzed by polyprenyl reductase Dfg10 together with some other unknown enzymes. The aim of this study was to identify such enzymes in Candida. The Dfg10 polyprenyl reductase from S. cerevisiae comprises a C-terminal 3-oxo-5-alpha-steroid 4-dehydrogenase domain. Alignment analysis revealed such a domain in two ORFs (orf19.209 and orf19.3293) from C. albicans, which were similar, respectively, to Dfg10 polyprenyl reductase and Tsc13 enoyl-transferase from S. cerevisiae. Deletion of orf19.209 in Candida impaired saturation of polyprenols. The Tsc13 homologue turned out not to be capable of saturating polyprenols, but limiting its expression reduce the cellular level of dolichols and polyprenols. This reduction was not due to a decreased expression of genes encoding cis-prenyltransferases from the dolichol branch but to a lower expression of genes encoding enzymes of the early stages of the mevalonate pathway. Despite the resulting lower consumption of acetyl-CoA, the sole precursor of the mevalonate pathway, it was not redirected towards fatty acid synthesis or elongation. Lowering the expression of TSC13 decreased the expression of the ACC1 gene encoding acetyl-CoA carboxylase, the key regulatory enzyme of fatty acid synthesis and elongation.     

Unprecedented diversity of catalytic domains in the first four modules of the putative pederin polyketide synthase

Polyketides of the pederin group are highly potent antitumor compounds found in terrestrial beetles and marine sponges. Pederin is used by beetles of the genera Paederus and Paederidus as a chemical defense. We have recently identified a group of putative pederin biosynthesis genes and localized them to the genome of an as yet unculturable Pseudomonas sp. symbiont, the likely true pederin producer. However, this polyketide synthase cluster lacks several genes expected for pederin production. Here we report an additional polyketide synthase encoded on a separate region of the symbiont genome. It contains at least three novel catalytic domains that are predicted to be involved in pederin chain initiation and the formation of an unusual exomethylene bond. The region is bordered by mobility pseudogenes; this suggests that gene transposition led to the disjointed cluster organization. With this work, all putative pederin genes have been identified. Their heterologous expression in a culturable bacterium will provide important insights into how sustainable sources of invertebrate-derived drug candidates can be created.     

Biosynthesis of the 2-pyridone tenellin in the insect pathogenic fungus Beauveria bassiana

Genomic DNA from the insect pathogenic fungus Beauveria bassiana was used as a template in a PCR with degenerate primers designed to amplify a fragment of a C-methyl transferase (CMeT) domain from a highly reduced fungal polyketide synthase (PKS). The resulting 270-bp PCR product was homologous to other fungal PKS CMeT domains and was used as a probe to isolate a 7.3-kb fragment of genomic DNA from a BamH1 library. Further library probing and TAIL-PCR then gave a 21.9-kb contig that encoded a 12.9-kb fused type I PKS-NRPS ORF together with ORFs encoding other oxidative and reductive enzymes. A directed knockout experiment with a BaR cassette, reported for the first time in B. bassiana, identified the PKS-NRPS as being involved in the biosynthesis of the 2-pyridone tenellin. Other fungal PKS-NRPS genes are known to be involved in the formation of tetramic acids in fungi, and it thus appears likely that related compounds are precursors of 2-pyridones in fungi. B. bassiana tenellin KO and WT strains proved to be equally pathogenic towards insect larvae; this indicated that tenellin is not involved in insect pathogenesis.     

Induction of Nectriapyrone Biosynthesis in the Rice Blast Fungus Pyricularia oryzae by Disturbance of the Two-Component Signal Transduction System

Most fungal secondary metabolism genes are poorly expressed under laboratory conditions. Nectriapyrones are known as secondary metabolites produced mainly by symbiotic fungi, including endophytes and plant pathogens. Herein, we show the induction of nectriapyrone production in the rice blast fungus Pyricularia oryzae. The two-component signal transduction system was disturbed by disrupting OSM1 and PoYPD1, which encoded a HOG MAP kinase and a His-containing phosphotransfer (HPt) protein, respectively. This induced the production of two polyketide compounds: nectriapyrone and its hydroxylated analogue. The nectriapyrone biosynthetic gene cluster consists of a polyketide synthase gene (NEC1) and an O-methyltransferase gene (NEC2). Overexpression of the two genes induced overproduction of nectriapyrone and five nectriapyrone analogues, including a new derivative. Nectriapyrone production was not required for the infection of rice. The structure of nectriapyrone is similar to that of the germicidins produced by Streptomyces spp., and nectriapyrone inhibited the growth of Streptomyces griseus.     

Biosynthetic Code for Divergolide Assembly in a Bacterial Mangrove Endophyte

Divergolides are structurally diverse ansamycins produced by a bacterial endophyte (Streptomyces sp.) of the mangrove tree Bruguiera gymnorrhiza. By genomic analyses a gene locus coding for the divergolide pathway was detected. The div gene cluster encodes genes for the biosynthesis of 3‐amino‐5‐hydroxybenzoate and the rare extender units ethylmalonyl‐CoA and isobutylmalonyl‐CoA, polyketide assembly by a modular type I polyketide synthase (PKS), and enzymes involved in tailoring reactions, such as a Baeyer–Villiger oxygenase. A detailed PKS domain analysis confirmed the stereochemical integrity of the divergolides and provided valuable new insights into the formation of the diverse aromatic chromophores. The bioinformatic analyses and the isolation and full structural elucidation of four new divergolide congeners led to a revised biosynthetic model that illustrates the formation of four different types of ansamycin chromophores from a single polyketide precursor.     

Ribosylhopane,a Novel Bacterial Hopanoid,as Precursor of C35 Bacteriohopanepolyols in Streptomyces coelicolor A3(2)

Wild‐type Streptomyces coelicolor A3(2) produces aminobacteriohopanetriol as the only elongated C₃₅ hopanoid. The hopanoid phenotype of two mutants bearing a deletion of genes from a previously identified hopanoid biosynthesis gene cluster provides clues to the formation of C₃₅ bacteriohopanepolyols. orf14 encodes a putative nucleosidase; its deletion induces the accumulation of adenosylhopane as it cannot be converted into ribosylhopane. orf18 encodes a putative transaminase; its deletion results in the accumulation of adenosylhopane, ribosylhopane, and bacteriohopanetetrol. Ribosylhopane was postulated twenty years ago as a precursor for bacterial hopanoids but was never identified in a bacterium. Absence of the transaminase encoded by orf18 prevents the reductive amination of ribosylhopane into aminobacteriohopanetriol and induces its accumulation. Its reduction by an aldose‐reductase‐like enzyme produces bacteriohopanetetrol, which is normally not present in S. coelicolor.     

Genome Mining of the Hitachimycin Biosynthetic Gene Cluster: Involvement of a Phenylalanine‐2,3‐aminomutase in Biosynthesis

Hitachimycin is a macrolactam antibiotic with (S)‐β‐phenylalanine (β‐Phe) at the starter position of its polyketide skeleton. To understand the incorporation mechanism of β‐Phe and the modification mechanism of the unique polyketide skeleton, the biosynthetic gene cluster for hitachimycin in Streptomyces scabrisporus was identified by genome mining. The identified gene cluster contains a putative phenylalanine‐2,3‐aminomutase (PAM), five polyketide synthases, four β‐amino‐acid‐carrying enzymes, and a characteristic amidohydrolase. A hitA knockout mutant showed no hitachimycin production, but antibiotic production was restored by feeding with (S)‐β‐Phe. We also confirmed the enzymatic activity of the HitA PAM. The results suggest that the identified gene cluster is responsible for the biosynthesis of hitachimycin. A plausible biosynthetic pathway for hitachimycin, including a unique polyketide skeletal transformation mechanism, is proposed.     

Identification of late-stage glycosylation steps in the biosynthetic pathway of the anthracycline nogalamycin

Nogalamycin is an anthracycline antibiotic that has been shown to exhibit significant cytotoxicity. Its biological activity requires two deoxysugar moieties: nogalose and nogalamine, which are attached at C7 and C1, respectively, of the aromatic polyketide aglycone. Curiously, the aminosugar nogalamine is also connected through a C-C bond between C2 and C5'. Despite extensive molecular genetic characterization of early biosynthetic steps, nogalamycin glycosylation has not been investigated in detail. Here we show that expression of the majority of the gene cluster in Streptomyces albus led to accumulation of three new anthracyclines, which unexpectedly included nogalamycin derivatives in which nogalamine was replaced either by rhodosamine with the C-C bond intact (nogalamycin R) or by 2-deoxyfucose without the C-C bond (nogalamycin F). In addition, a monoglycosylated intermediate-3',4'-demethoxynogalose-1-hydroxynogalamycinone-was isolated. Importantly, when the remaining biosynthetic genes were introduced into the heterologous host by using a two-plasmid system, nogalamycin could be isolated from the cultures, thus indicating that the whole gene cluster had been identified. We further show that one of the three glycosyltransferases (GTs) residing in the cluster-snogZ-appears to be redundant, whereas gene inactivation experiments revealed that snogE and snogD act as nogalose and nogalamine transferases, respectively. The substrate specificity of the nogalamine transferase SnogD was demonstrated in vitro: the enzyme was able to remove 2deoxyfucose from nogalamycin F. All of the new compounds were found to inhibit human topoisomerase I in activity measurements, whereas only nogalamycin R showed minor activity against topoisomerase II.     

Non-colinear polyketide biosynthesis in the aureothin and neoaureothin pathways: an evolutionary perspective

Aureothin and neoaureothin (spectinabilin) represent rare nitroaryl-substituted polyketide metabolites from Streptomyces thioluteus and Streptomyces orinoci, respectively, which only differ in the lengths of the polyene backbones. Cloning and sequencing of the 39 kb neoaureothin (nor) biosynthesis gene cluster and its comparison with the aureothin (aur) pathway genes revealed that both polyketide synthase (PKS) assembly lines are remarkably similar. In both cases the module architecture breaks with the principle of colinearity, as individual PKS modules are used in an iterative fashion. Parsimony and neighbour-joining phylogenetic studies provided insights into the evolutionary process that led to the programming of these unusual type I PKS systems and to prediction of which modules act iteratively. The iterative function of the first module in the neoaureothin pathway, NorA, was confirmed by a successful cross-complementation.     

Secondary Metabolites of the Rice Blast Fungus Pyricularia oryzae: Biosynthesis and Biological Function

Plant pathogenic fungi produce a wide variety of secondary metabolites with unique and complex structures. However, most fungal secondary metabolism genes are poorly expressed under laboratory conditions. Moreover, the relationship between pathogenicity and secondary metabolites remains unclear. To activate silent gene clusters in fungi, successful approaches such as epigenetic control, promoter exchange, and heterologous expression have been reported. Pyricularia oryzae, a well-characterized plant pathogenic fungus, is the causal pathogen of rice blast disease. P. oryzae is also rich in secondary metabolism genes. However, biosynthetic genes for only four groups of secondary metabolites have been well characterized in this fungus. Biosynthetic genes for two of the four groups of secondary metabolites have been identified by activating secondary metabolism. This review focuses on the biosynthesis and roles of the four groups of secondary metabolites produced by P. oryzae. These secondary metabolites include melanin, a polyketide compound required for rice infection; pyriculols, phytotoxic polyketide compounds; nectriapyrones, antibacterial polyketide compounds produced mainly by symbiotic fungi including endophytes and plant pathogens; and tenuazonic acid, a well-known mycotoxin produced by various plant pathogenic fungi and biosynthesized by a unique NRPS-PKS enzyme.     

Artificial reconstruction of two cryptic angucycline antibiotic biosynthetic pathways

Genome-sequencing projects have revealed that Streptomyces bacteria have the genetic potential to produce considerably larger numbers of natural products than can be observed under standard laboratory conditions. Cryptic angucycline-type aromatic polyketide gene clusters are particularly abundant. Sequencing of two such clusters from Streptomyces sp. PGA64 and H021 revealed the presence of several open reading frames that could be involved in processing the basic angucyclic carbon skeleton. The pga gene cluster contains one putative FAD-dependant monooxygenase (pgaE) and a putatively bifunctional monooxygenase/short chain alcohol reductase (pgaM), whereas the cab cluster contains two similar monooxygenases (cabE and cabM) and an independent reductase (cabV). In this study we have reconstructed the biosynthetic pathways for aglycone synthesis by cloning and sequentially expressing the angucycline tailoring genes with genes required for the synthesis of the unmodified angucycline metabolite-UWM6-in Streptomyces lividans TK24. The expression studies unequivocally showed that, after the production of UWM6, the pathways proceed through the action of the similar monooxygenases PgaE and CabE, followed by reactions catalysed by PgaM and CabMV. Analysis of the metabolites produced revealed that addition of pgaE and cabE genes directs both pathways to a known shunt product, rabelomycin, whereas expression of all genes from a given pathway results in the production of the novel angucycline metabolites gaudimycin A and B. However, one of the end products is most probably further modified by endogenous S. lividans TK24 enzymes. These experiments demonstrate that genes that are either inactive or cryptic in their native host can be used as biosynthetic tools to generate new compounds.     

Catalytic relationships between type I and type II iterative polyketide synthases: The Aspergillus parasiticus norsolorinic acid synthase

Norsolorinic acid synthase (NSAS) is a type I iterative polyketide synthase that occurs in the filamentous fungus Aspergillus parasiticus. PCR was used to clone fragments of NSAS corresponding to the acyl carrier protein (ACP), acyl transferase (AT) and beta-ketoacyl-ACP synthase (KS) catalytic domains. Expression of these gene fragments in Escherichia coli led to the production of soluble ACP and AT proteins. Coexpression of ACP with E. coli holo-ACP synthase (ACPS) let to production of NSAS holo-ACP, which could also be formed in vitro by using Streptomyces coelicolor ACPS. Analysis by mass spectrometry showed that, as with other type I carrier proteins, self-malonylation is not observed in the presence of malonyl CoA alone. However, the NSAS holo-ACP serves as substrate for S. coelicolor MCAT, S. coelicolor actinorhodin holo-ACP and NSAS AT domain-catalysed malonate transfer from malonyl CoA. The AT domain could transfer malonate from malonyl CoA to NSAS holo-ACP, but not hexanoate or acetate from either the cognate CoA or FAS ACP species to NSAS holo-ACP. The NSAS holo-ACP was also active in actinorhodin minimal PKS assays, but only in the presence of exogenous malonyl transferases.     

Discovery and Elucidation of the Biosynthesis of Aspernidgulenes: Novel Polyenes from Aspergillus Nidulans by Using Serial Promoter Replacement

Through serial promoter exchanges, we isolated several novel polyenes, the aspernidgulenes, from Aspergillus nidulans and uncovered their succinct biosynthetic pathway involving only four enzymes. An enoyl reductase (ER)-less highly reducing polyketide synthase (HR-PKS) putatively produces a 5,6-dihydro-α-pyrone polyene, which undergoes bisepoxidation, epoxide ring opening, cyclization, and hydrolytic cleavage by three tailoring enzymes to generate aspernidgulene A1 and A2. Our findings demonstrate the prowess of fungal-tailoring enzymes to transform a polyketide scaffold concisely and efficiently into complex structures. Moreover, comparison with citreoviridin and aurovertin biosynthesis suggests that methylation of the α-pyrone hydroxy group by methyltransferase (CtvB or AurB) is the branching point at which the biosynthesis of these two classes of compounds diverge. Therefore, scanning for the presence or absence of the gatekeeping α-pyrone methyltransferase gene in homologous clusters might be a potential way to classify the product bioinformatically as belonging to methylated α-pyrone polyenes or polyenes containing rings derived from the cyclization of the unmethylated 5,6-dihydro-α-pyrone, such as 2,3-dimethyl-γ-lactone and oxabicyclo[2.2.1]heptane.