Tilmann Weber

antiSMASH 3.0—a comprehensive resource for the genome mining of biosynthetic gene clusters

Abstract Microbial secondary metabolism constitutes a rich source of antibiotics, chemotherapeutics, insecticides and other high-value chemicals. Genome mining of gene clusters that encode the biosynthetic pathways for these metabolites has become a key methodology for novel compound discovery. In 2011, we introduced antiSMASH, a web server and stand-alone tool for the automatic genomic identification and analysis of biosynthetic gene clusters, available at http://antismash.secondarymetabolites.org. Here, we present version 3.0 of antiSMASH, which has undergone major improvements. A full integration of the recently published ClusterFinder algorithm now allows using this probabilistic algorithm to detect putative gene clusters of unknown types. Also, a new dereplication variant of the ClusterBlast module now identifies similarities of identified clusters to any of 1172 clusters with known end products. At the enzyme level, active sites of key biosynthetic enzymes are now pinpointed through a curated pattern-matching procedure and Enzyme Commission numbers are assigned to functionally classify all enzyme-coding genes. Additionally, chemical structure prediction has been improved by incorporating polyketide reduction states. Finally, in order for users to be able to organize and analyze multiple antiSMASH outputs in a private setting, a new XML output module allows offline editing of antiSMASH annotations within the Geneious software.

antiSMASH: rapid identification, annotation and analysis of secondary metabolite biosynthesis gene clusters in bacterial and fungal genome sequences

Bacterial and fungal secondary metabolism is a rich source of novel bioactive compounds with potential pharmaceutical applications as antibiotics, anti-tumor drugs or cholesterol-lowering drugs. To find new drug candidates, microbiologists are increasingly relying on sequencing genomes of a wide variety of microbes. However, rapidly and reliably pinpointing all the potential gene clusters for secondary metabolites in dozens of newly sequenced genomes has been extremely challenging, due to their biochemical heterogeneity, the presence of unknown enzymes and the dispersed nature of the necessary specialized bioinformatics tools and resources. Here, we present antiSMASH (antibiotics & Secondary Metabolite Analysis Shell), the first comprehensive pipeline capable of identifying biosynthetic loci covering the whole range of known secondary metabolite compound classes (polyketides, non-ribosomal peptides, terpenes, aminoglycosides, aminocoumarins, indolocarbazoles, lantibiotics, bacteriocins, nucleosides, beta-lactams, butyrolactones, siderophores, melanins and others). It aligns the identified regions at the gene cluster level to their nearest relatives from a database containing all other known gene clusters, and integrates or cross-links all previously available secondary-metabolite specific gene analysis methods in one interactive view. antiSMASH is available at http://antismash.secondarymetabolites.org.

antiSMASH 2.0—a versatile platform for genome mining of secondary metabolite producers

Microbial secondary metabolites are a potent source of antibiotics and other pharmaceuticals. Genome mining of their biosynthetic gene clusters has become a key method to accelerate their identification and characterization. In 2011, we developed antiSMASH, a web-based analysis platform that automates this process. Here, we present the highly improved antiSMASH 2.0 release, available at http://antismash.secondarymetabolites.org/. For the new version, antiSMASH was entirely re-designed using a plug-and-play concept that allows easy integration of novel predictor or output modules. antiSMASH 2.0 now supports input of multiple related sequences simultaneously (multi-FASTA/GenBank/EMBL), which allows the analysis of draft genomes comprising multiple contigs. Moreover, direct analysis of protein sequences is now possible. antiSMASH 2.0 has also been equipped with the capacity to detect additional classes of secondary metabolites, including oligosaccharide antibiotics, phenazines, thiopeptides, homo-serine lactones, phosphonates and furans. The algorithm for predicting the core structure of the cluster end product is now also covering lantipeptides, in addition to polyketides and non-ribosomal peptides. The antiSMASH ClusterBlast functionality has been extended to identify sub-clusters involved in the biosynthesis of specific chemical building blocks. The new features currently make antiSMASH 2.0 the most comprehensive resource for identifying and analyzing novel secondary metabolite biosynthetic pathways in microorganisms.

Specificity prediction of adenylation domains in nonribosomal peptide synthetases (NRPS) using transductive support vector machines (TSVMs)

We present a new support vector machine (SVM)-based approach to predict the substrate specificity of subtypes of a given protein sequence family. We demonstrate the usefulness of this method on the example of aryl acid-activating and amino acid-activating adenylation domains (A domains) of nonribosomal peptide synthetases (NRPS). The residues of gramicidin synthetase A that are 8 Å around the substrate amino acid and corresponding positions of other adenylation domain sequences with 397 known and unknown specificities were extracted and used to encode this physico-chemical fingerprint into normalized real-valued feature vectors based on the physico-chemical properties of the amino acids. The SVM software package SVMlight was used for training and classification, with transductive SVMs to take advantage of the information inherent in unlabeled data. Specificities for very similar substrates that frequently show cross-specificities were pooled to the so-called composite specificities and predictive models were built for them. The reliability of the models was confirmed in cross-validations and in comparison with a currently used sequence-comparison-based method. When comparing the predictions for 1230 NRPS A domains that are currently detectable in UniProt, the new method was able to give a specificity prediction in an additional 18% of the cases compared with the old method. For 70% of the sequences both methods agreed, for <6% they did not, mainly on low-confidence predictions by the existing method. None of the predictive methods could infer any specificity for 2.4% of the sequences, suggesting completely new types of specificity.

Phylogenetic analysis of condensation domains in NRPS sheds light on their functional evolution.

BackgroundNon-ribosomal peptide synthetases (NRPSs) are large multimodular enzymes that synthesize a wide range of biologically active natural peptide compounds, of which many are pharmacologically important. Peptide bond formation is catalyzed by the Condensation (C) domain. Various functional subtypes of the C domain exist: An LCL domain catalyzes a peptide bond between two L-amino acids, a DCL domain links an L-amino acid to a growing peptide ending with a D-amino acid, a Starter C domain (first denominated and classified as a separate subtype here) acylates the first amino acid with a β-hydroxy-carboxylic acid (typically a β-hydroxyl fatty acid), and Heterocyclization (Cyc) domains catalyze both peptide bond formation and subsequent cyclization of cysteine, serine or threonine residues. The homologous Epimerization (E) domain flips the chirality of the last amino acid in the growing peptide; Dual E/C domains catalyze both epimerization and condensation.ResultsIn this paper, we report on the reconstruction of the phylogenetic relationship of NRPS C domain subtypes and analyze in detail the sequence motifs of recently discovered subtypes (Dual E/C, DCL and Starter domains) and their characteristic sequence differences, mutually and in comparison with LCL domains. Based on their phylogeny and the comparison of their sequence motifs, LCL and Starter domains appear to be more closely related to each other than to other subtypes, though pronounced differences in some segments of the protein account for the unequal donor substrates (amino vs. β-hydroxy-carboxylic acid). Furthermore, on the basis of phylogeny and the comparison of sequence motifs, we conclude that Dual E/C and DCL domains share a common ancestor. In the same way, the evolutionary origin of a C domain of unknown function in glycopeptide (GP) NRPSs can be determined to be an LCL domain. In the case of two GP C domains which are most similar to DCL but which have LCL activity, we postulate convergent evolution.ConclusionWe systematize all C domain subtypes including the novel Starter C domain. With our results, it will be easier to decide the subtype of unknown C domains as we provide profile Hidden Markov Models (pHMMs) for the sequence motifs as well as for the entire sequences. The determined specificity conferring positions will be helpful for the mutation of one subtype into another, e.g. turning DCL to LCL, which can be a useful step for obtaining novel products.

antiSMASH 4.0-improvements in chemistry prediction and gene cluster boundary identification

Abstract Many antibiotics, chemotherapeutics, crop protection agents and food preservatives originate from molecules produced by bacteria, fungi or plants. In recent years, genome mining methodologies have been widely adopted to identify and characterize the biosynthetic gene clusters encoding the production of such compounds. Since 2011, the ‘antibiotics and secondary metabolite analysis shell—antiSMASH’ has assisted researchers in efficiently performing this, both as a web server and a standalone tool. Here, we present the thoroughly updated antiSMASH version 4, which adds several novel features, including prediction of gene cluster boundaries using the ClusterFinder method or the newly integrated CASSIS algorithm, improved substrate specificity prediction for non-ribosomal peptide synthetase adenylation domains based on the new SANDPUMA algorithm, improved predictions for terpene and ribosomally synthesized and post-translationally modified peptides cluster products, reporting of sequence similarity to proteins encoded in experimentally characterized gene clusters on a per-protein basis and a domain-level alignment tool for comparative analysis of trans-AT polyketide synthase assembly line architectures. Additionally, several usability features have been updated and improved. Together, these improvements make antiSMASH up-to-date with the latest developments in natural product research and will further facilitate computational genome mining for the discovery of novel bioactive molecules.

CRISPR-Cas9 Based Engineering of Actinomycetal Genomes

Bacteria of the order Actinomycetales are one of the most important sources of pharmacologically active and industrially relevant secondary metabolites. Unfortunately, many of them are still recalcitrant to genetic manipulation, which is a bottleneck for systematic metabolic engineering. To facilitate the genetic manipulation of actinomycetes, we developed a highly efficient CRISPR-Cas9 system to delete gene(s) or gene cluster(s), implement precise gene replacements, and reversibly control gene expression in actinomycetes. We demonstrate our system by targeting two genes, actIORF1 (SCO5087) and actVB (SCO5092), from the actinorhodin biosynthetic gene cluster in Streptomyces coelicolor A3(2). Our CRISPR-Cas9 system successfully inactivated the targeted genes. When no templates for homology-directed repair (HDR) were present, the site-specific DNA double-strand breaks (DSBs) introduced by Cas9 were repaired through the error-prone nonhomologous end joining (NHEJ) pathway, resulting in a library of deletions with variable sizes around the targeted sequence. If templates for HDR were provided at the same time, precise deletions of the targeted gene were observed with near 100% frequency. Moreover, we developed a system to efficiently and reversibly control expression of target genes, deemed CRISPRi, based on a catalytically dead variant of Cas9 (dCas9). The CRISPR-Cas9 based system described here comprises a powerful and broadly applicable set of tools to manipulate actinomycetal genomes.

Metabolic engineering of antibiotic factories: new tools for antibiotic production in actinomycetes

Actinomycetes are excellent sources for novel bioactive compounds, which serve as potential drug candidates for antibiotics development. While industrial efforts to find and develop novel antimicrobials have been severely reduced during the past two decades, the increasing threat of multidrug-resistant pathogens and the development of new technologies to find and produce such compounds have again attracted interest in this field. Based on improvements in whole-genome sequencing, novel methods have been developed to identify the secondary metabolite biosynthetic gene clusters by genome mining, to clone them, and to express them in heterologous hosts in much higher throughput than before. These technologies now enable metabolic engineering approaches to optimize production yields and to directly manipulate the pathways to generate modified products.

Molecular Analysis of the Kirromycin Biosynthetic Gene Cluster Revealed β-Alanine as Precursor of the Pyridone Moiety

Kirromycin is a complex linear polyketide that acts as a protein biosynthesis inhibitor by binding to the bacterial elongation factor Tu. The kirromycin biosynthetic gene cluster was isolated from the producer, Streptomyces collinus Tü 365, and confirmed by targeted disruption of essential biosynthesis genes. Kirromycin is synthesized by a large hybrid polyketide synthase (PKS)/nonribosomal peptide synthetase (NRPS) encoded by the genes kirAI-kirAVI. This complex involves some very unusual features, including the absence of internal acyltransferase (AT) domains in KirAI-KirAV, multiple split-ups of PKS modules on separate genes, and swapping in the domain organization. Interestingly, one PKS enzyme, KirAVI, contains internal AT domains. Based on in silico analysis, a route to pyridone formation involving PKS and NRPS steps was postulated. This hypothesis was experimentally proven by feeding studies with [U-13C3(15)N]beta-alanine and NMR and MS analyses of the isolated pure kirromycin.

Comparative analysis and insights into the evolution of gene clusters for glycopeptide antibiotic biosynthesis.

The bal, cep, dbv, sta and tcp gene clusters specify the biosynthesis of the glycopeptide antibiotics balhimycin, chloroeremomycin, A40926, A47934 and teicoplanin, respectively. These structurally related compounds share a similar mechanism of action in their inhibition of bacterial cell wall formation. Comparative sequence analysis was performed on the five gene clusters. Extensive conserved synteny was observed between the bal and cep clusters, which direct the synthesis of very similar compounds but originate from two different species of the genus Amycolatopsis. All other cluster pairs show a limited degree of conserved synteny, involving biosynthetically functional gene cassettes: these include those involved in the synthesis of the carbon backbone of two non-proteinogenic amino acids; in the linkage of amino acids 1–3 and 4–7 in the heptapeptide; and in the formation of the aromatic cross-links. Furthermore, these segments of conserved synteny are often preceded by conserved intergenic regions. Phylogenetic analysis of protein families shows several instances in which relatedness in the chemical structure of the glycopeptides is not reflected in the extent of the relationship of the corresponding polypeptides. Coherent branchings are observed for all polypeptides encoded by the syntenous gene cassettes. These results suggest that the acquisition of distinct, functional genetic elements has played a significant role in the evolution of glycopeptide gene clusters, giving them a mosaic structure. In addition, the synthesis of the structurally similar compounds A40926 and teicoplanin appears as the result of convergent evolution.