Enzymatic assembly of epothilones: the EpoC subunit and reconstitution of the EpoA-ACP/B/C polyketide and nonribosomal peptide interfaces

The biosynthesis of epothilones, a family of hybrid polyketide (PK)/nonribosomal peptide (NRP) antitumor agents, provides an ideal system to study a hybrid PK/NRP natural product with significant biomedical value. Here the third enzyme involved in epothilone production, the five domain 195 kDa polyketide synthase (PKS) EpoC protein, has been expressed and purified from Escherichia coli. EpoC was combined with the first two enzymes of the epothilone biosynthesis pathway, the acyl carrier protein (ACP) domain of EpoA and EpoB, to reconstitute the early steps in epothilone biosynthesis. The acyltransferase (AT) domain of EpoC transfers the methylmalonyl moiety from methylmalonyl-CoA to the holo HS-acyl carrier protein (ACP) in an autoacylation reaction. The ketosynthase (KS) domain of EpoC decarboxylates the methylmalonyl-S-EpoC acyl enzyme to generate the carbon nucleophile that reacts with methylthiazolylcarboxyl-S-EpoB. The resulting condensation product can be reduced in the presence of NADPH by the ketoreductase (KR) domain of EpoC and then dehydrated by the dehydratase (DH) domain to produce the methylthiazolylmethylacrylyl-S-EpoC acyl enzyme intermediate that serves as the acyl donor for subsequent elongation of the epothilone chain. The acetyl-CoA donor can be replaced with propionyl-CoA, isobutyryl-CoA, and benzoyl-CoA and the acyl chains accepted by both EpoB and EpoC subunits to produce ethyl-, isopropyl-, and phenylthiazolylmethylacrylyl-S-EpoC acyl enzyme intermediates, suggesting that future combinatorial biosynthetic variations in epothilone assembly may be feasible. These results demonstrate in vitro reconstitution of both the PKS/NRPS interface (EpoA-ACP/B) and the NRPS/PKS interface (EpoB/C) in the assembly line for this antitumor natural product.     

Excision of the epothilone synthetase B cyclization domain and demonstration of in trans condensation/cyclodehydration activity

The epothilones are potent anticancer natural products produced by a polyketide synthase (PKS)-nonribosomal peptide synthetase (NRPS) hybrid involving proteins EpoA-F. The single NRPS module of the epothilone assembly line, EpoB, is a distinct subunit of approximately 160 kDa and consists of four successive domains: cyclization, adenylation, oxidation, and peptidyl carrier protein (Cy-A-Ox-PCP). The cyclization domain is responsible for introduction of the thiazoline heterocycle into the growing polyketide/nonribosomal peptide chain from the precursors malonyl-CoA and cysteine through the multiple steps of condensation, cyclization, and dehydration. This enzyme-bound thiazoline intermediate is subsequently oxidized to a thiazole by the EpoB Ox domain. The EpoB module was dissected to provide 57 kDa EpoB(Cy) and 102 kDa EpoB(A-Ox-PCP) as subunit fragments to evaluate Cy as a free-standing domain. EpoB was reconstituted by these fragments in trans to generate the methylthiazole product. Using this system, apparent kinetic constants for the upstream acyl donor EpoA(ACP) and EpoB(Cy) were determined, providing a measure of affinity for the naturally occurring interface of the amino terminus of EpoB and the EpoA carboxy terminus. Site-directed mutants in excised EpoB(Cy) were prepared and used to examine residues involved in condensation and heterocycle formation. This work demonstrates the ability to define a functional Cy domain by excision from its native NRPS module, and examine both its protein-protein interactions and mechanism of activity.     

Isolation and characterization of the epothilone biosynthetic gene cluster from Sorangium cellulosum

The epothilone biosynthetic gene cluster was isolated from Sorangium cellulosum strain SMP44. The gene cluster contains seven genes and spans approx. 56kb. The genes encoding the PKS, epoA, epoC, epoD, epoE, and epoF, are divided into nine modules. The EpoB protein is a non-ribosomal peptide synthetase (NRPS) that catalyzes formation of the thiazole found in the epothilones. EpoK is a P450 enzyme responsible for the epoxidation of epothilones C and D to epothilones A and B, respectively. EpoK was expressed in Escherichia coli, and the purified protein was shown to convert epothilone D to epothilone B in vitro.     

Heterologous production of epothilones B and D in Streptomyces venezuelae

Epothilones, produced from the myxobacterium Sorangium cellulosum, are potential anticancer agents that stabilize microtubules in a similar manner to paclitaxel. The entire epothilone biosynthetic gene cluster was heterologously expressed in an engineered strain of Streptomyces venezuelae bearing a deletion of pikromycin polyketide synthase gene cluster. The resulting strains produced approximately 0.1 μg/l of epothilone B as a sole product after 4 days cultivation. Deletion of an epoF encoding the cytochrome P450 epoxidase gave rise to a mutant that selectively produces 0.4 μg/l of epothilone D. To increase the production level of epothilones B and D, an additional copy of the positive regulatory gene pikD was introduced into the chromosome of both S. venezuleae mutant strains. The resulting strains showed enhanced production of corresponding compounds (approximately 2-fold). However, deletion of putative transport genes, orf3 and orf14 in the epothilone D producing S. venezuelae mutant strain, led to an approximately 3-fold reduction in epothilone D production. These results introduce S. venezuelae as an alternative heterologous host for the production of these valuable anticancer agents and demonstrate the possibility of engineering this strain as a generic heterologous host for the production of polyketides and hybrid polyketide-nonribosomal peptides.     

Redesign, synthesis and functional expression of the 6-deoxyerythronolide B polyketide synthase gene cluster

A generic design of Type I polyketide synthase genes has been reported in which modules, and domains within modules, are flanked by sets of unique restriction sites that are repeated in every module [1]. Using the universal design, we synthesized the six-module DEBS gene cluster optimized for codon usage in E. coli, and cloned the three open reading frames into three compatible expression vectors. With one correctable exception, the amino acid substitutions required for restriction site placements were compatible with polyketide production. When expressed in E. coli the codon-optimized synthetic gene cluster produced significantly more protein than did the wild-type sequence. Indeed, for optimal polyketide production, PKS expression had to be down-regulated by promoter attenuation to achieve balance with expression of the accessory proteins needed to support polyketide biosynthesis.     

Oxidase domains in epothilone and bleomycin biosynthesis: thiazoline to thiazole oxidation during chain elongation

The natural products epothilone and bleomycin are assembled by hybrid polyketide/nonribosomal peptide synthetases. Of note in these assembly lines is the conversion of internal cysteine residues into thiazolines and their subsequent oxidation to heteroaromatic thiazole rings. We have excised the EpoB oxidase domain, EpoB-Ox, proposed to be responsible for thiazoline to thiazole oxidation in epothilone biosynthesis, and expressed it in soluble form in Escherichia coli. The purified domain is an FMN-containing flavoprotein that demonstrates thiazoline to thiazole oxidase activity when incubated with thioester substrate mimics. Kinetic parameters were determined for both thiazoline and oxazoline substrates, with k(cat) values ranging between 48.8 and 0.55 min(-1). While the physiological electron acceptor is not yet known, molecular oxygen is needed in these in vitro assays to mediate reoxidation of reduced FMN. Additionally, the oxidase domain-containing BlmIII from the bleomycin assembly line was heterologously expressed and purified. BlmIII is also an FMN-containing protein with activity similar to EpoB-Ox. This work marks the first direct characterization of nonribosomal peptide synthetase oxidase domain activity and will lead to further exploration of these flavoproteins.     

Diversity of epothilone producers among Sorangium strains in producer‐positive soil habitats

Large‐scale surveys show that the anti‐tumour compounds known as epothilones are produced by only a small proportion of Sorangium strains, thereby greatly hampering the research and development of these valuable compounds. In this study, to investigate the niche diversity of epothilone‐producing Sorangium strains, we re‐surveyed four soil samples where epothilone producers were previously found. Compared with the < 2.5% positive strains collected from different places, epothilone producers comprised 25.0–75.0% of the Sorangium isolates in these four positive soil samples. These sympatric epothilone producers differed not only in their 16S rRNA gene sequences and morphologies but also in their production of epothilones and biosynthesis genes. A further exploration of 14 soil samples collected from a larger area around a positive site showed a similar high positive ratio of epothilone producers among the Sorangium isolates. The present results suggest that, in an area containing epothilone producers, the long‐term genetic variations and refinements resulting from selective pressure form a large reservoir of epothilone‐producing Sorangium strains with diverse genetic compositions.     

Heterologous production of epothilone C and D in Escherichia coli

The epothilones are a family of polyketide natural products that show a high potential as anticancer drugs. They are synthesized by the action of a hybrid nonribosomal peptide synthetase/polyketide synthase in the myxobacterium Sorangium cellulosum. In this work, the genes encoding the entire cluster,epoA, epoB, epoC, epoD, epoE, and epoF, were redesigned and synthesized to allow for expression in Escherichia coli. The expression of the largest of the proteins, EpoD, also required the protein be separated into two polypeptides with compatible module linkers. Using a combination of lowered temperature, chaperone coexpression, and alternative promoters, we succeeded in producing a soluble protein from all genes in the epothilone cluster. The entire synthetic epothilone cluster was then expressed in a strain of E. coli modified to enable polyketide biosynthesis, resulting in the production of epothilones C and D. Furthermore, feeding a thioester of the normal substrate for EpoD to cells expressing the epoD, epoE, and epoF genes also led to the production of epothilones C and D. The design of the synthetic epothilone genes together with E. coli expression provides the ideal platform for both the biochemical investigation of the epothilone PKS and the generation of novel biosynthetic epothilone analogues.     

The interaction between mitotic checkpoint proteins,CENP-E and BubR1, is diminished in epothilone B-resistant A549 cells

Centromere associated protein-E (CENP-E), a mitotic checkpoint protein, is required for efficient, stable microtubule capture at kinetochores during mitosis. Absence of CENP-E results in misaligned chromosomes leading to metaphase arrest. Microtubule-interacting agents such as Taxol and epothilone B (EpoB), at concentrations that induce mitotic arrest, transiently increase expression of CENP-E in a variety of cancer cell lines. The CENP-E level in an EpoB-resistant A549 cell line, EpoB40, is ~ 2-fold higher than in A549 cells. CENP-E overexpression, after transfection with CENP-E cDNA into drug sensitive cells, does not alter Taxol or EpoB sensitivity. However, suppression of CENP-E expression by CENP-E siRNA results in a moderate increase in drug sensitivity, suggesting that a minimal quantity of CENP-E is required for maintaining its function. It is known that CENP-E binds to BubR1 and enhances its recruitment to each unattached kinetochore. Suppression of CENP-E results in a decrease in BubR1 levels in EpoB40 cells. During metaphase, both targeting of CENP-E and BubR1 to the kinetochores and the interaction between CENP-E and BubR1 are significantly reduced in EpoB40 cells, compared to A549 cells. In addition, the distance between the two centrosomes during metaphase is shorter in EpoB40 than in A549 cells, suggesting that defects in the spindle-assembly checkpoint have occurred in EpoB40 cells during the development of drug resistance. These results indicate that defects in the mitotic checkpoint may have a role in, or be the result of, the development of EpoB resistance.     

Naturally mosaic operons for secondary metabolite biosynthesis: variability and putative horizontal transfer of discrete catalytic domains of the epothilone polyketide synthase locus

A putative instance of horizontal gene transfer (HGT) involving adjacent, discrete -ketoacyl synthase (KS), acyl carrier protein (ACP) and nonribosomal peptide synthase (NRPS) domains of the epothilone Type I polyketide biosynthetic gene cluster from the myxobacterium Sorangium cellulosom was identified using molecular phylogenetics and sequence analyses. The specific KS domain of the module EPO B fails to cluster phylogenetically with other epothilone KS sequences present at this locus, in contrast to what is typically observed in many other Type I polyketide synthase (PKS) biosynthetic loci. Furthermore, the GC content of the epoB KS, epoA ACP and NRPS domains differs significantly from the base composition of other epothilone domain sequences. In addition, the putatively transferred epothilone loci are located near previously identified transposon-like sequences. Lastly, comparison with other KS loci revealed another possible case of horizontal transfer of secondary metabolite genes in the genus Pseudomonas. This study emphasizes the use of several lines of concordant evidence (phylogenetics, base composition, transposon sequences) to infer the evolutionary history of particular gene and enzyme sequences, and the results support the idea that genes coding for adaptive traits, e.g. defensive natural products, may be prone to transposition between divergent prokaryotic taxa and genomes.Communicated by W. Arber     

埃博霉素（Epothilones）的PKS/NRPS杂合基因簇

埃博霉素是由粘细菌纤维堆囊菌产生的一类具有促微管聚合活性的大环内酯类化合物。埃博霉素生物合成的多酶复合体是一个由多个功能模块组成,同时含有多聚酮合酶（PKS）和非核糖体肽合成酶（NRPS）的大操纵子。根据同位素标记试验结果和合成酶全基因簇功能的推测,埃博霉素的生物合成包括聚酮链的引发、链合成的起始和噻唑环的形成、链的延伸和转移、链合成的终止释放和环化、及产物的后修饰5个阶段。埃博霉素的PKS/NRPS杂合基因簇是开展组合生物合成研究的良好材料。     

Modulation of epothilone analog production through media design

Recently, the epothilone polyketide synthase (PKS) was successfully introduced into a heterologous production host for the large-scale production of epothilone D. We have found that at least three other epothilones can also be produced as the major fermentation product of this recombinant strain by supplementation of specific substrates to the production media. Addition of acetate or propionate to the media results in modulation of the epothilone D:C ratio, whereas addition of l-serine with either acetate or propionate yields epothilone H₁ or H₂ as the major product. This strategy permits production of at least four novel epothilones by culturing a single host with a genetically modified epothilone PKS in various media. Journal of Industrial Microbiology & Biotechnology (2002) 28, 17–20 DOI: 10.1038/sj/jim/7000209 Received 20 June 2001/ Accepted in revised form 03 September 2001     

Glycosylation and production characteristics of epothilones in alkali-tolerant <Emphasis Type="Italic">Sorangium cellulosum</Emphasis> strain So0157-2

3-O-α-D-ribofuranosyl epothilone A (epothiloneoside A) is a major component of glycosylated epothilones in Sorangium cellulosum strain So0157-2. The production and glycosylation ratios of epothiloneoside A in both solid and liquid culture conditions with various pH values and carbon sources were studied. The results showed that glycosylation occurs whenever epothilones are produced, regardless of changes in pH values, production time curves, and different carbon sources. We suggest that glycosylation is a stable process, paralleling the biosynthesis of epothilones in the So0157-2 strain.     

Sequence analysis and gene amplification study of the penicillin biosynthesis gene cluster from different strains of <Emphasis Type="Italic">Penicillium chrysogenum</Emphasis>

Industrial strains of Penicillium chrysogenum possess many genomic changes leading to higher levels of penicillin. In this work several production and wild-type strains of Penicillium chrysogenum were used in comparative nucleotide sequence analysis of the biosynthesis cluster. The alignments confirmed sequence conservation not only in promoter regions of the biosynthesis genes but also throughout the entire 44.7-kbp genomic fragment comprising the whole biosynthesis cluster with 15.5-kbp and 13.1-kbp flanking regions. As another titre-enhancing mechanism we subsequently examined gene dosage in two production strains introduced here, NMU2/40 and B14. Quantitative real-time PCR and Southern blot analysis showed the amplification of the biosynthesis genes in both these strains. Through the real-time PCR method the exact copy number was estimated for each of the pcbAB, pcbC and penDE genes. The equal pool of all three genes per genome was confirmed for the both production strains indicating that in these strains the entire penicillin cluster has been amplified as an intact element. Penicillium chrysogenum NMU2/40 was found to carry four copies of the cluster, while six copies were estimated for B14. This also proves the contribution of the additional titre-enhancing mechanisms in both strains, since the industrial data referred much higher production of these strains compared with the single copy reference strain NRRL 1951.     

Understanding the genetics of regulation of aflatoxin production and Aspergillus flavus development

Aflatoxins are polyketide-derived, toxic, and carcinogenic secondary metabolites produced primarily by two fungal species, Aspergillus flavus and A. parasiticus, on crops such as corn, peanuts, cottonseed, and treenuts. Regulatory guidelines issued by the U.S. Food and Drug Administration (FDA) prevent sale of commodities if contamination by these toxins exceeds certain levels. The biosynthesis of these toxins has been extensively studied. About 15 stable precursors have been identified. The genes involved in encoding the proteins required for the oxidative and regulatory steps in the biosynthesis are clustered in a 70 kb portion of chromosome 3 in the A. flavus genome. With the characterization of the gene cluster, new insights into the cellular processes that govern the genes involved in aflatoxin biosynthesis have been revealed, but the signaling processes that turn on aflatoxin biosynthesis during fungal contamination of crops are still not well understood. New molecular technologies, such as gene microarray analyses, quantitative polymerase chain reaction (PCR), and chromatin immunoprecipitation are being used to understand how physiological stress, environmental and soil conditions, receptivity of the plant, and fungal virulence lead to episodic outbreaks of aflatoxin contamination in certain commercially important crops. With this fundamental understanding, we will be better able to design improved non-aflatoxigenic biocompetitive Aspergillus strains and develop inhibitors of aflatoxin production (native to affected crops or otherwise) amenable to agricultural application for enhancing host-resistance against fungal invasion or toxin production. Comparisons of aflatoxin-producing species with other fungal species that retain some of the genes required for aflatoxin formation is expected to provide insight into the evolution of the aflatoxin gene cluster, and its role in fungal physiology. Therefore, information on how and why the fungus makes the toxin will be valuable for developing an effective and lasting strategy for control of aflatoxin contamination.     

Cloning and characterization of the gene cluster required for beauvericin biosynthesis in Fusarium proliferatum

Beauvericin, a cyclohexadepsipeptide-possessing natural product with synergistic antifungal, insecticidal, and cytotoxic activities. We isolated and characterized the fpBeas gene cluster, devoted to beauvericin biosynthesis, from the filamentous fungus Fusarium proliferatum LF061. Targeted inactivation of the F. proliferatum genomic copy of fpBeas abolished the production of beauvericin. Comparative sequence analysis of the FpBEAS showed 74% similarity with the BbBEAS that synthesizes the cyclic trimeric ester beauvericin in Beauveria bassiana, which assembles N-methyl-dipeptidol monomer intermediates by the programmed iterative use of the nonribosomal peptide synthetase modules. Differences between the organization of the beauvericin loci in F. proliferaturm and B. bassiana revealed the mechanism for high production of beauvericin in F. proliferatum. Our work provides new insights into beauvericin biosynthesis, and may lead to beauvericin overproduction and creation of new analogs via synthetic biology approaches.     

Recombinant production of docosahexaenoic acid in a polyketide biosynthesis mode in Escherichia coli

The docosahexaenoic acid (DHA) biosynthesis gene cluster (pDHA3) from the DHA-producing Moritella marina strain MP-1 includes the genes pfaA, pfaB, pfaC, and pfaD, which are similar to the genes of polyketide biosynthesis. When this cluster was co-expressed in Escherichia coli with M. marina MP-1 pfaE, which encodes phosphopantetheinyl transferase, DHA was biosynthesized. The maximum production of DHA (5% of total fatty acids) was observed at 15°C. This is the first report of the recombinant production of DHA in a polyketide biosynthesis mode.     

Reconstitution of a mini-gene cluster combined with ribosome engineering led to effective enhancement of salinomycin production in Streptomyces albus

Salinomycin, an FDA-approved polyketide drug, was recently identified as a promising anti-tumour and anti-viral lead compound. It is produced by Streptomyces albus, and the biosynthetic gene cluster (sal) spans over 100 kb. The genetic manipulation of large polyketide gene clusters is challenging, and approaches delivering reliable efficiency and accuracy are desired. Herein, a delicate strategy to enhance salinomycin production was devised and evaluated. We reconstructed a minimized sal gene cluster (mini-cluster) on pSET152 including key genes responsible for tailoring modification, antibiotic resistance, positive regulation and precursor supply. These genes were overexpressed under the control of constitutive promoter P_kasO* or P_neo. The pks operon was not included in the mini-cluster, but it was upregulated by SalJ activation. After the plasmid pSET152::mini-cluster was introduced into the wild-type strain and a chassis host strain obtained by ribosome engineering, salinomycin production was increased to 2.3-fold and 5.1-fold compared with that of the wild-type strain respectively. Intriguingly, mini-cluster introduction resulted in much higher production than overexpression of the whole sal gene cluster. The findings demonstrated that reconstitution of sal mini-cluster combined with ribosome engineering is an efficient novel approach and may be extended to other large polyketide biosynthesis.     

Evolution of the aflatoxin gene cluster

WhyAspergillus species produce aflatoxin remains an unsolved question. In this report we suggest that evolution of the aflatoxin biosynthesis gene cluster has been a multistep process. More than 300 million years ago a primordial cluster of genes allowed production of anthraquinones that may have served as insect attractants to facilitate spore dispersal. Later adaptive evolutionary steps introduced genes into the cluster that encoded enzymes associated with fungal virulence. These genes may have allowed the otherwise saprophytic fungi to be better able to colonize living plants. Later, genes for production of aflatoxins B1 and G1 were added to the basal cluster. Loss of the ability to produce aflatoxin G1 occurred with the divergence ofA. flavus, a species that, perhaps, was more successful than its ancestors at colonizing plants. This logical progression in evolutionary development of the aflatoxin biosynthetic cluster fits the phylogenetic data as well as known chemical reactivity of the initially formed anthraquinone polyketide metabolites.     

Improved production of A40926 by Nonomuraea sp. through deletion of a pathway-specific acetyltransferase

Nonomuraea strain ATCC 39727 produces the glycopeptide A40926, used for manufacturing dalbavancin, currently in advanced clinical trials. From the gene cluster involved in A40926 biosynthesis, a strain deleted in dbv23 was constructed. This mutant can produce only the glycopeptides lacking the O-linked acetyl residue at position 6 of the mannose moiety, while, under identical fermentation conditions, the wild-type strain produces mostly glycopeptides carrying an acetylated mannose. Furthermore, the total amount of glycopeptides produced by the mutant strain was found to be approximately twice that of the wild type. The reduced level of glycopeptides observed in the wild-type strain may be due to an inhibitory effect exerted by the acetylated compound on the biosynthesis of A40926. Indeed, spiking production cultures with ≥1 μg/ml of the acetylated glycopeptide inhibited A40926 production in the mutant strain.