Computational approach for prediction of domain organization and substrate specificity of modular polyketide synthases

Modular polyketide synthases (PKSs) are large multi-enzymatic, multi-domain megasynthases, which are involved in the biosynthesis of a class of pharmaceutically important natural products, namely polyketides. These enzymes harbor a set of repetitive active sites termed modules and the domains present in each module dictate the chemical moiety that would add to a growing polyketide chain. This modular logic of biosynthesis has been exploited with reasonable success to produce several novel compounds by genetic manipulation. However, for harnessing their vast potential of combinatorial biosynthesis, it is essential to develop knowledge based in silico approaches for correlating the sequence and domain organization of PKSs to their polyketide products. In this work, we have carried out extensive sequence analysis of experimentally characterized PKS clusters to develop an automated computational protocol for unambiguous identification of various PKS domains in a polypeptide sequence. A structure based approach has been used to identify the putative active site residues of acyltransferase (AT) domains, which control the specificities for various starter and extender units during polyketide biosynthesis. On the basis of the analysis of the active site residues and molecular modelling of substrates in the active site of representative AT domains, we have identified a crucial residue that is likely to play a major role in discriminating between malonate and methylmalonate during selection of extender groups by this domain. Structural modelling has also explained the experimentally observed chiral preference of AT domain in substrate selection. This computational protocol has been used to predict the domain organization and substrate specificity for PKS clusters from various microbial genomes. The results of our analysis as well as the computational tools for prediction of domain organization and substrate specificity have been organized in the form of a searchable computerized database (PKSDB). PKSDB would serve as a valuable tool for identification of polyketide products biosynthesized by uncharacterized PKS clusters. This database can also provide guidelines for rational design of experiments to engineer novel polyketides.     

Towards Prediction of Metabolic Products of Polyketide Synthases: An In Silico Analysis

Sequence data arising from an increasing number of partial and complete genome projects is revealing the presence of the polyketide synthase (PKS) family of genes not only in microbes and fungi but also in plants and other eukaryotes. PKSs are huge multifunctional megasynthases that use a variety of biosynthetic paradigms to generate enormously diverse arrays of polyketide products that posses several pharmaceutically important properties. The remarkable conservation of these gene clusters across organisms offers abundant scope for obtaining novel insights into PKS biosynthetic code by computational analysis. We have carried out a comprehensive in silico analysis of modular and iterative gene clusters to test whether chemical structures of the secondary metabolites can be predicted from PKS protein sequences. Here, we report the success of our method and demonstrate the feasibility of deciphering the putative metabolic products of uncharacterized PKS clusters found in newly sequenced genomes. Profile Hidden Markov Model analysis has revealed distinct sequence features that can distinguish modular PKS proteins from their iterative counterparts. For iterative PKS proteins, structural models of iterative ketosynthase (KS) domains have revealed novel correlations between the size of the polyketide products and volume of the active site pocket. Furthermore, we have identified key residues in the substrate binding pocket that control the number of chain extensions in iterative PKSs. For modular PKS proteins, we describe for the first time an automated method based on crucial intermolecular contacts that can distinguish the correct biosynthetic order of substrate channeling from a large number of non-cognate combinatorial possibilities. Taken together, our in silico analysis provides valuable clues for formulating rules for predicting polyketide products of iterative as well as modular PKS clusters. These results have promising potential for discovery of novel natural products by genome mining and rational design of novel natural products.     

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.     

Metagenomic analysis reveals diverse polyketide synthase gene clusters in microorganisms associated with the marine sponge Discodermia dissoluta

Sponge-associated bacteria are thought to produce many novel bioactive compounds, including polyketides. PCR amplification of ketosynthase domains of type I modular polyketide synthases (PKS) from the microbial community of the marine sponge Discodermia dissoluta revealed great diversity and a novel group of sponge-specific PKS ketosynthase domains. Metagenomic libraries totaling more than four gigabases of bacterial genomes associated with this sponge were screened for type I modular PKS gene clusters. More than 90% of the clones in total sponge DNA libraries represented bacterial DNA inserts, and 0.7% harbored PKS genes. The majority of the PKS hybridizing clones carried small PKS clusters of one to three modules, although some clones encoded large multimodular PKSs (more than five modules). The most abundant large modular PKS appeared to be encoded by a bacterial symbiont that made up < 1% of the sponge community. Sequencing of this PKS revealed 14 modules that, if expressed and active, is predicted to produce a multimethyl-branched fatty acid reminiscent of mycobacterial lipid components. Metagenomic libraries made from fractions enriched for unicellular or filamentous bacteria differed significantly, with the latter containing numerous nonribosomal peptide synthetase (NRPS) and mixed NRPS-PKS gene clusters. The filamentous bacterial community of D. dissoluta consists mainly of Entotheonella spp., an unculturable sponge-specific taxon previously implicated in the biosynthesis of bioactive peptides.     

Developing tools for engineering hybrid polyketide synthetic pathways

Bacterial type I polyketide synthases (PKSs) are complex, multifunctional enzymes that synthesize structurally diverse and medicinally important natural products. Given their modular organization, the manipulation of type I PKSs holds tremendous promise for the generation of novel compounds that are not easily accessible by standard synthetic chemical approaches. In theory, hybrid polyketide synthetic pathways can be constructed through the rational recombination of catalytic domains or modules from a variety of PKS systems; however, the general success of this strategy has been elusive, largely due to a poor understanding of the interactions between catalytic domains, as well as PKS modules. Over the past several years, a fundamental knowledge of these issues, and others, has begun to emerge, offering refined strategies for the facile engineering of hybrid polyketide pathways.     

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.     

Metagenomic Analysis Reveals Diverse Polyketide Synthase Gene Clusters in Microorganisms Associated with the Marine Sponge Discodermia dissoluta

Sponge-associated bacteria are thought to produce many novel bioactive compounds, including polyketides. PCR amplification of ketosynthase domains of type I modular polyketide synthases (PKS) from the microbial community of the marine sponge Discodermia dissoluta revealed great diversity and a novel group of sponge-specific PKS ketosynthase domains. Metagenomic libraries totaling more than four gigabases of bacterial genomes associated with this sponge were screened for type I modular PKS gene clusters. More than 90% of the clones in total sponge DNA libraries represented bacterial DNA inserts, and 0.7% harbored PKS genes. The majority of the PKS hybridizing clones carried small PKS clusters of one to three modules, although some clones encoded large multimodular PKSs (more than five modules). The most abundant large modular PKS appeared to be encoded by a bacterial symbiont that made up <1% of the sponge community. Sequencing of this PKS revealed 14 modules that, if expressed and active, is predicted to produce a multimethyl-branched fatty acid reminiscent of mycobacterial lipid components. Metagenomic libraries made from fractions enriched for unicellular or filamentous bacteria differed significantly, with the latter containing numerous nonribosomal peptide synthetase (NRPS) and mixed NRPS-PKS gene clusters. The filamentous bacterial community of D. dissoluta consists mainly of Entotheonella spp., an unculturable sponge-specific taxon previously implicated in the biosynthesis of bioactive peptides.     

A novel docking domain interface model predicting recombination between homoeologous modular biosynthetic gene clusters

An in silico model for homoeologous recombination between gene clusters encoding modular polyketide synthases (PKS) or non-ribosomal peptide synthetases (NRPS) was developed. This model was used to analyze recombination between 12 PKS clusters from Streptomyces species and related genera to predict if new clusters might give rise to new products. In many cases, there were only a limited number of recombination sites (about 13 per cluster pair), suggesting that recombination may pose constraints on the evolution of PKS clusters. Most recombination events occurred between pairs of ketosynthase (KS) domains, allowing the biosynthetic outcome of the recombinant modules to be predicted. About 30% of recombinants were predicted to produce polyketides. Four NRPS clusters from Streptomyces strains were also used for in silico recombination. They yielded a comparable number of recombinants to PKS clusters, but the adenylation (A) domains contained the largest proportion of recombination events; this might be a mechanism for producing new substrate specificities. The extreme G + C-content, the presence of linear chromosomes and plasmids, as well as the lack of a mutSL-mismatch repair system should favor production of recombinants in Streptomyces species.     

Isolation and characterization of a reducing polyketide synthase gene from the lichen-forming fungus Usnea longissima

The reducing polyketide synthases found in filamentous fungi are involved in the biosynthesis of many drugs and toxins. Lichens produce bioactive polyketides, but the roles of reducing polyketide synthases in lichens remain to be clearly elucidated. In this study, a reducing polyketide synthase gene (U1PKS3) was isolated and characterized from a cultured mycobiont of Usnea longissima. Complete sequence information regarding U1PKS3 (6,519 bp) was obtained by screening a fosmid genomic library. A U1PKS3 sequence analysis suggested that it contains features of a reducing fungal type I polyketide synthase with β-ketoacyl synthase (KS), acyltransferase (AT), dehydratase (DH), enoyl reductase (ER), ketoacyl reducatse (KR), and acyl carrier protein (ACP) domains. This domain structure was similar to the structure of ccRadsl, which is known to be involved in resorcylic acid lactone biosynthesis in Chaetomium chiversii. The results of phylogenetic analysis located U1PKS3 in the clade of reducing polyketide synthases. RT-PCR analysis results demonstrated that UIPKS3 had six intervening introns and that UIPKS3 expression was upregulated by glucose, sorbitol, inositol, and mannitol.     

Active-site residue,domain and module swaps in modular polyketide synthases

Sequence comparisons of multiple acyltransferase (AT) domains from modular polyketide synthases (PKSs) have highlighted a correlation between a short sequence motif and the nature of the extender unit selected. When this motif was specifically altered in the bimodular model PKS DEBS1-TE of Saccharopolyspora erythraea, the products included triketide lactones in which acetate extension units had been incorporated instead of propionate units at the predicted positions. We also describe a cassette system for convenient construction of hybrid modular PKSs based on the tylosin PKS in Streptomyces fradiae and demonstrate its use in domain and module swaps.     

Genetic construction and functional analysis of hybrid polyketide synthases containing heterologous acyl carrier proteins.

The gene that encodes the acyl carrier protein (ACP) of the actinorhodin polyketide synthase (PKS) of Streptomyces coelicolor A3(2) was replaced with homologs from the granaticin, oxytetracycline, tetracenomycin, and putative frenolicin polyketide synthase gene clusters. All of the replacements led to expression of functional synthases, and the recombinants synthesized aromatic polyketides similar in chromatographic properties to actinorhodin or to shunt products produced by mutants defective in the actinorhodin pathway. Some regions within the ACP were also shown to be interchangeable and allow production of a functional hybrid ACP. Structural analysis of the most abundant polyketide product of one of the recombinants by electrospray mass spectrometry suggested that it is identical to mutactin, a previously characterized shunt product of an actVII mutant (deficient in cyclase and dehydrase activities). Quantitative differences in the product profiles of strains that express the various hybrid synthases were observed. These can be explained, at least in part, by differences in ribosome-binding sites upstream of each ACP gene, implying either that the ACP concentration in some strains is rate limiting to overall PKS activity or that the level of ACP expression also influences the expression of another enzyme(s) encoded by a downstream gene(s) in the same operon as the actinorhodin ACP gene. These results reaffirm the idea that construction of hybrid polyketide synthases will be a useful approach for dissecting the molecular basis of the specificity of PKS-catalyzed reactions. However, they also point to the need for reducing the chemical complexity of the approach by minimizing the diversity of polyketide products synthesized in strains that produce recombinant polyketide synthases.     

Characterization of the biosynthetic gene cluster for the antifungal polyketide soraphen A from Sorangium cellulosum So ce26

A genomic DNA region of over 80 kb that contains the complete biosynthetic gene cluster for the synthesis of the antifungal polyketide metabolite soraphen A was cloned from Sorangium cellulosum So ce26. The nucleotide sequence of the soraphen A gene region, including 67,523 bp was determined. Examination of this sequence led to the identification of two adjacent type I polyketide synthase (PKS) genes that encode the soraphen synthase. One of the soraphen A PKS genes includes three biosynthetic modules and the second contains five additional modules for a total of eight. The predicted substrate specificities of the acyltransferase (AT) domains, as well as the reductive loop domains identified within each module, are consistent with expectations from the structure of soraphen A. Genes were identified in the regions flanking the two soraphen synthase genes that are proposed to have roles in the biosynthesis of soraphen A. Downstream of the soraphen PKS genes is an O-methyltransferase (OMT) gene. Upstream of the soraphen PKS genes there is a gene encoding a reductase and a group of genes that are postulated to have roles in the synthesis of methoxymalonyl-acyl carrier protein (ACP). This unusual extender unit is proposed to be incorporated in two positions of the soraphen polyketide chain. One of the genes in this group contains distinct domains for an AT, an ACP, and an OMT.     

Evolution of metabolic diversity: Insights from microbial polyketide synthases

Polyketides are a family of complex natural products that are built from simple carboxylic acid building blocks. In microorganisms, the majority of these secondary metabolites are produced by exceptionally large, multifunctional proteins termed polyketide synthases (PKSs). Each unit of a type I PKS assembly line resembles a mammalian type fatty acid synthase (FAS), although certain domains are optionally missing. The evolutionary analysis of microbial PKS has revealed a long joint evolution process of PKSs and FASs. The phylogenomic analysis of modular type I PKSs as the most widespread PKS type in bacteria showed a large impact of gene duplications and gene losses on the evolution of type I PKS in different bacterial groups. The majority of type I PKSs in actinobacteria and cyanobacteria may have evolved from a common ancestor, whereas in proteobacteria most type I PKSs were acquired from other bacterial groups. The modularization of type I PKSs almost unexceptionally started with multiple duplications of a single ancestor module. The repeating modules represent ideal platforms for recombination events that can lead to corresponding changes in the actual chemistry of the products. The analysis of these “natural reprogramming” events of PKSs may assist in the development of concepts for the biocombinatorial design of bioactive compounds.     

An approach for obtaining perfect hybridization probes for unknown polyketide synthase genes: a search for the epothilone gene cluster

An approach is described for obtaining 'perfect probes' for type I modular polyketide synthase (PKS) gene clusters that in turn enables the identification of all such gene clusters in a genome. The approach involves sequencing small fragments of a random genomic DNA library containing one or more modular PKS gene clusters, and identifying which fragments emanate from PKS genes. Knowing the approximate sizes of the genome and the target gene cluster, one can predict the the frequency that a PKS gene fragment will be present in the library sequenced. Computer simulations of the approach were applied to the known PKS and non-ribosomal peptide synthetase (NRPS) gene clusters in the Bacillus subtilus genome. The approach was then used to identify PKS gene fragments in a strain of Sorangium cellulosum that produces epothilone. In addition to identifying fragments of the epothilone gene cluster, we obtained 11 unique fragments from other PKS gene clusters; the results suggest that there may be six to eight PKS gene clusters in this organism. In addition, we identified four unique fragments of NRPS genes, demonstrating that the approach is also applicable for identification of these modular gene clusters.     

Horizontal gene transfer and gene conversion drive evolution of modular polyketide synthases

Soil bacteria live in a very competitive environment and produce many secondary metabolites; there appears to be strong selective pressure for evolution of new compounds. Secondary metabolites are the most important source of chemical structures for the pharmaceutical industry and an understanding of the evolutionary process should help in finding novel chemical entities. Modular polyketide synthases are a particularly interesting case for evolutionary studies, because much of the chemical structure can be predicted from DNA sequence. Previous evolutionary studies have concentrated on individual modules or domains and were not able to study the evolution of orthologues. This study overcame this problem by considering complete clusters as "organisms", so that orthologous modules and domains could be identified and used to characterise evolutionary pathways. Seventeen modular polyketide synthase clusters were identified that fell into six classes. Gene conversion within clusters was very common (affecting about 15?% of domains) and was detected by discordance in phylogenetic trees. An evolutionary model is proposed in which a single cross over between two different clusters (i.e. horizontal gene transfer) would generate a cluster of very different architecture with radically different chemical products; subsequent gene conversion and deletions would explore chemical variants. Two probable examples of such recombination were found. This model suggests strategies for detecting horizontal gene transfer in cluster evolution.     

Solution structure and proposed domain domain recognition interface of an acyl carrier protein domain from a modular polyketide synthase

Polyketides are a medicinally important class of natural products. The architecture of modular polyketide synthases (PKSs), composed of multiple covalently linked domains grouped into modules, provides an attractive framework for engineering novel polyketide-producing assemblies. However, impaired domain-domain interactions can compromise the efficiency of engineered polyketide biosynthesis. To facilitate the study of these domain-domain interactions, we have used nuclear magnetic resonance (NMR) spectroscopy to determine the first solution structure of an acyl carrier protein (ACP) domain from a modular PKS, 6-deoxyerythronolide B synthase (DEBS). The tertiary fold of this 10-kD domain is a three-helical bundle; an additional short helix in the second loop also contributes to the core helical packing. Superposition of residues 14-94 of the ensemble on the mean structure yields an average atomic RMSD of 0.64 +/- 0.09 Angstrom for the backbone atoms (1.21 +/- 0.13 Angstrom for all non-hydrogen atoms). The three major helices superimpose with a backbone RMSD of 0.48 +/- 0.10 Angstrom (0.99 +/- 0.11 Angstrom for non-hydrogen atoms). Based on this solution structure, homology models were constructed for five other DEBS ACP domains. Comparison of their steric and electrostatic surfaces at the putative interaction interface (centered on helix II) suggests a model for protein-protein recognition of ACP domains, consistent with the previously observed specificity. Site-directed mutagenesis experiments indicate that two of the identified residues influence the specificity of ACP recognition.