Bioprospecting for Genes Encoding Hydrocarbon-Degrading Enzymes from Metagenomic Samples Isolated from Northern Adriatic Sea Sediments.

Three metagenomic libraries were constructed using surface sediment samples from the northern Adriatic Sea. Two of the samples were taken from a highly polluted and an unpolluted site respectively. The third sample from a polluted site had been enriched using crude oil. The results of the metagenome analyses were incorporated in the REDPET relational database (http://redpet.bioinfo.pbf.hr/REDPET), which was generated using the previously developed MEGGASENSE platform. The database includes taxonomic data to allow the assessment of the biodiversity of metagenomic libraries and a general functional analysis of genes using hidden Markov model (HMM) profiles based on the KEGG database. A set of 22 specialised HMM profiles was developed to detect putative genes for hydrocarbon-degrading enzymes. Use of these profiles showed that the metagenomic library generated after selection on crude oil had enriched genes for aerobic n-alkane degradation. The use of this system for bioprospecting was exemplified using potential alkB and almA genes from this library.

Proteomics links the redox state to calcium signaling during bleaching of the scleractinian coral Acropora microphthalma on exposure to high solar irradiance and thermal stress.

Shipboard experiments were each performed over a 2 day period to examine the proteomic response of the symbiotic coral Acropora microphthalma exposed to acute conditions of high temperature/low light or high light/low temperature stress. During these treatments, corals had noticeably bleached. The photosynthetic performance of residual algal endosymbionts was severely impaired but showed signs of recovery in both treatments by the end of the second day. Changes in the coral proteome were determined daily and, using recently available annotated genome sequences, the individual contributions of the coral host and algal endosymbionts could be extracted from these data. Quantitative changes in proteins relevant to redox state and calcium metabolism are presented. Notably, expression of common antioxidant proteins was not detected from the coral host but present in the algal endosymbiont proteome. Possible roles for elevated carbonic anhydrase in the coral host are considered: to restore intracellular pH diminished by loss of photosynthetic activity, to indirectly limit intracellular calcium influx linked with enhanced calmodulin expression to impede late-stage symbiont exocytosis, or to enhance inorganic carbon transport to improve the photosynthetic performance of algal symbionts that remain in hospite. Protein effectors of calcium-dependent exocytosis were present in both symbiotic partners. No caspase-family proteins associated with host cell apoptosis, with exception of the autophagy chaperone HSP70, were detected, suggesting that algal loss and photosynthetic dysfunction under these experimental conditions were not due to host-mediated phytosymbiont destruction. Instead, bleaching occurred by symbiont exocytosis and loss of light-harvesting pigments of algae that remain in hospite. These proteomic data are, therefore, consistent with our premise that coral endosymbionts can mediate their own retention or departure from the coral host, which may manifest as "symbiont shuffling" of Symbiodinium clades in response to environmental stress.

MEGGASENSE - The Metagenome/Genome Annotated Sequence Natural Language Search Engine: A Platform for the Construction of Sequence Data Warehouses.

The MEGGASENSE platform constructs relational databases of DNA or protein sequences. The default functional analysis uses 14 106 hidden Markov model (HMM) profiles based on sequences in the KEGG database. The Solr search engine allows sophisticated queries and a BLAST search function is also incorporated. These standard capabilities were used to generate the SCATT database from the predicted proteome of Streptomyces cattleya. The implementation of a specialised metagenome database (AMYLOMICS) for bioprospecting of carbohydrate-modifying enzymes is described. In addition to standard assembly of reads, a novel 'functional' assembly was developed, in which screening of reads with the HMM profiles occurs before the assembly. The AMYLOMICS database incorporates additional HMM profiles for carbohydrate-modifying enzymes and it is illustrated how the combination of HMM and BLAST analyses helps identify interesting genes. A variety of different proteome and metagenome databases have been generated by MEGGASENSE.

Gene duplications are extensive and contribute significantly to the toxic proteome of nematocysts isolated from Acropora digitifera (Cnidaria: Anthozoa: Scleractinia).

BACKGROUND: Gene duplication followed by adaptive selection is a well-accepted process leading to toxin diversification in venoms. However, emergent genomic, transcriptomic and proteomic evidence now challenges this role to be at best equivocal to other processess . Cnidaria are arguably the most ancient phylum of the extant metazoa that are venomous and such provide a definitive ancestral anchor to examine the evolution of this trait. METHODS: Here we compare predicted toxins from the translated genome of the coral Acropora digitifera to putative toxins revealed by proteomic analysis of soluble proteins discharged from nematocysts, to determine the extent to which gene duplications contribute to venom innovation in this reef-building coral species. A new bioinformatics tool called HHCompare was developed to detect potential gene duplications in the genomic data, which is made freely available ( https://github.com/rgacesa/HHCompare ). RESULTS: A total of 55 potential toxin encoding genes could be predicted from the A. digitifera genome, of which 36 (65 %) had likely arisen by gene duplication as evinced using the HHCompare tool and verified using two standard phylogeny methods. Surprisingly, only 22 % (12/55) of the potential toxin repertoire could be detected following rigorous proteomic analysis, for which only half (6/12) of the toxin proteome could be accounted for as peptides encoded by the gene duplicates. Biological activities of these toxins are dominatedby putative phospholipases and toxic peptidases. CONCLUSIONS: Gene expansions in A. digitifera venom are the most extensive yet described in any venomous animal, and gene duplication plays a significant role leading to toxin diversification in this coral species. Since such low numbers of toxins were detected in the proteome, it is unlikely that the venom is evolving rapidly by prey-driven positive natural selection. Rather we contend that the venom has a defensive role deterring predation or harm from interspecific competition and overgrowth by fouling organisms. Factors influencing translation of toxin encoding genes perhaps warrants more profound experimental consideration.

Predicting substrate specificity of adenylation domains of nonribosomal peptide synthetases and other protein properties by latent semantic indexing.

Successful genome mining is dependent on accurate prediction of protein function from sequence. This often involves dividing protein families into functional subtypes (e.g., with different substrates). In many cases, there are only a small number of known functional subtypes, but in the case of the adenylation domains of nonribosomal peptide synthetases (NRPS), there are >500 known substrates. Latent semantic indexing (LSI) was originally developed for text processing but has also been used to assign proteins to families. Proteins are treated as ''documents'' and it is necessary to encode properties of the amino acid sequence as ''terms'' in order to construct a term-document matrix, which counts the terms in each document. This matrix is then processed to produce a document-concept matrix, where each protein is represented as a row vector. A standard measure of the closeness of vectors to each other (cosines of the angle between them) provides a measure of protein similarity. Previous work encoded proteins as oligopeptide terms, i.e. counted oligopeptides, but used no information regarding location of oligopeptides in the proteins. A novel tokenization method was developed to analyze information from multiple alignments. LSI successfully distinguished between two functional subtypes in five well-characterized families. Visualization of different ''concept'' dimensions allows exploration of the structure of protein families. LSI was also used to predict the amino acid substrate of adenylation domains of NRPS. Better results were obtained when selected residues from multiple alignments were used rather than the total sequence of the adenylation domains. Using ten residues from the substrate binding pocket performed better than using 34 residues within 8 Å of the active site. Prediction efficiency was somewhat better than that of the best published method using a support vector machine.

Evolutionary concepts in natural products discovery: what actinomycetes have taught us.

Actinomycetes are a very important source of natural products for the pharmaceutical industry and other applications. Most of the strains belong to Streptomyces or related genera, partly because they are particularly amenable to growth in the laboratory and industrial fermenters. It is unlikely that chemical synthesis can fulfil the needs of the pharmaceutical industry for novel compounds so there is a continuing need to find novel natural products. An evolutionary perspective can help this process in several ways. Genome mining attempts to identify secondary metabolite biosynthetic clusters in DNA sequences, which are likely to produce interesting chemical entities. There are often technical problems in assembling the DNA sequences of large modular clusters in genome and metagenome projects, which can be overcome partially using information about the evolution of the domain sequences. Understanding the evolutionary mechanisms of modular clusters should allow simulation of evolutionary pathways in the laboratory to generate novel compounds.

Oxytetracycline hyper-production through targeted genome reduction of Streptomyces rimosus.

Most biosynthetic gene clusters (BGC) encoding the synthesis of important microbial secondary metabolites, such as antibiotics, are either silent or poorly expressed; therefore, to ensure a strong pipeline of novel antibiotics, there is a need to develop rapid and efficient strain development approaches. This study uses comparative genome analysis to instruct rational strain improvement, using Streptomyces rimosus, the producer of the important antibiotic oxytetracycline (OTC) as a model system. Sequencing of the genomes of two industrial strains M4018 and R6-500, developed independently from a common ancestor, identified large DNA rearrangements located at the chromosome end. We evaluated the effect of these genome deletions on the parental S. rimosus Type Strain (ATCC 10970) genome where introduction of a 145 kb deletion close to the OTC BGC in the Type Strain resulted in massive OTC overproduction, achieving titers that were equivalent to M4018 and R6-500. Transcriptome data supported the hypothesis that the reason for such an increase in OTC biosynthesis was due to enhanced transcription of the OTC BGC and not due to enhanced substrate supply. We also observed changes in the expression of other cryptic BGCs; some metabolites, undetectable in ATCC 10970, were now produced at high titers. This study demonstrated for the first time that the main force behind BGC overexpression is genome rearrangement. This new approach demonstrates great potential to activate cryptic gene clusters of yet unexplored natural products of medical and industrial value.IMPORTANCEThere is a critical need to develop novel antibiotics to combat antimicrobial resistance. Streptomyces species are very rich source of antibiotics, typically encoding 20-60 biosynthetic gene clusters (BGCs). However, under laboratory conditions, most are either silent or poorly expressed so that their products are only detectable at nanogram quantities, which hampers drug development efforts. To address this subject, we used comparative genome analysis of industrial Streptomyces rimosus strains producing high titers of a broad spectrum antibiotic oxytetracycline (OTC), developed during decades of industrial strain improvement. Interestingly, large-scale chromosomal deletions were observed. Based on this information, we carried out targeted genome deletions in the native strain S. rimosus ATCC 10970, and we show that a targeted deletion in the vicinity of the OTC BGC significantly induced expression of the OTC BGC, as well as some other silent BGCs, thus suggesting that this approach may be a useful way to identify new natural products.

KEGG orthology-based annotation of the predicted proteome of Acropora digitifera: ZoophyteBase - an open access and searchable database of a coral genome.

BACKGROUND: Contemporary coral reef research has firmly established that a genomic approach is urgently needed to better understand the effects of anthropogenic environmental stress and global climate change on coral holobiont interactions. Here we present KEGG orthology-based annotation of the complete genome sequence of the scleractinian coral Acropora digitifera and provide the first comprehensive view of the genome of a reef-building coral by applying advanced bioinformatics. DESCRIPTION: Sequences from the KEGG database of protein function were used to construct hidden Markov models. These models were used to search the predicted proteome of A. digitifera to establish complete genomic annotation. The annotated dataset is published in ZoophyteBase, an open access format with different options for searching the data. A particularly useful feature is the ability to use a Google-like search engine that links query words to protein attributes. We present features of the annotation that underpin the molecular structure of key processes of coral physiology that include (1) regulatory proteins of symbiosis, (2) planula and early developmental proteins, (3) neural messengers, receptors and sensory proteins, (4) calcification and Ca2+-signalling proteins, (5) plant-derived proteins, (6) proteins of nitrogen metabolism, (7) DNA repair proteins, (8) stress response proteins, (9) antioxidant and redox-protective proteins, (10) proteins of cellular apoptosis, (11) microbial symbioses and pathogenicity proteins, (12) proteins of viral pathogenicity, (13) toxins and venom, (14) proteins of the chemical defensome and (15) coral epigenetics. CONCLUSIONS: We advocate that providing annotation in an open-access searchable database available to the public domain will give an unprecedented foundation to interrogate the fundamental molecular structure and interactions of coral symbiosis and allow critical questions to be addressed at the genomic level based on combined aspects of evolutionary, developmental, metabolic, and environmental perspectives.

Databases of the thiotemplate modular systems (CSDB) and their in silico recombinants (r-CSDB).

Modular biosynthetic clusters are responsible for the synthesis of many important pharmaceutical products. They include polyketide synthases (PKS clusters), non-ribosomal synthetases (NRPS clusters), and mixed clusters (containing both PKS and NRPS modules). The ClustScan database (CSDB) contains highly annotated descriptions of 170 clusters. The database has a hierarchical organization, which allows easy extraction of DNA and protein sequences of polypeptides, modules, and domains as well as an organization of the annotation so as to be able to predict the product chemistry to view it or export it in a standard SMILES format. The recombinant ClustScan database contains information about predicted recombinants between PKS clusters. The recombinants are generated by modeling homologous recombination and are associated with annotation and prediction of product chemistry automatically generated by the model. The database contains over 20,000 recombinants and is a resource for in silico approaches to detecting promising new compounds. Methods are available to construct the corresponding recombinants in the laboratory.

Annotation of the modular polyketide synthase and nonribosomal peptide synthetase gene clusters in the genome of Streptomyces tsukubaensis NRRL18488.

The high G+C content and large genome size make the sequencing and assembly of Streptomyces genomes more difficult than for other bacteria. Many pharmaceutically important natural products are synthesized by modular polyketide synthases (PKSs) and nonribosomal peptide synthetases (NRPSs). The analysis of such gene clusters is difficult if the genome sequence is not of the highest quality, because clusters can be distributed over several contigs, and sequencing errors can introduce apparent frameshifts into the large PKS and NRPS proteins. An additional problem is that the modular nature of the clusters results in the presence of imperfect repeats, which may cause assembly errors. The genome sequence of Streptomyces tsukubaensis NRRL18488 was scanned for potential PKS and NRPS modular clusters. A phylogenetic approach was used to identify multiple contigs belonging to the same cluster. Four PKS clusters and six NRPS clusters were identified. Contigs containing cluster sequences were analyzed in detail by using the ClustScan program, which suggested the order and orientation of the contigs. The sequencing of the appropriate PCR products confirmed the ordering and allowed the correction of apparent frameshifts resulting from sequencing errors. The product chemistry of such correctly assembled clusters could also be predicted. The analysis of one PKS cluster showed that it should produce a bafilomycin-like compound, and reverse transcription (RT)-PCR was used to show that the cluster was transcribed.

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.

Recombinatorial biosynthesis of polyketides.

Modular polyketide synthases (PKSs) from Streptomyces and related genera of bacteria produce many important pharmaceuticals. A program called CompGen was developed to carry out in silico homologous recombination between gene clusters encoding PKSs and determine whether recombinants have cluster architectures compatible with the production of polyketides. The chemical structure of recombinant polyketides was also predicted. In silico recombination was carried out for 47 well-characterised clusters. The predicted recombinants would produce 11,796 different polyketide structures. The molecular weights and average degree of reduction of the chemical structures are dispersed around the parental structures indicating that they are likely to include pharmaceutically interesting compounds. The details of the recombinants and the chemical structures were entered in a database called r-CSDB. The virtual compound library is a useful resource for computer-aided drug design and chemoinformatics strategies for finding pharmaceutically relevant chemical entities. A strategy to construct recombinant Streptomyces strains to produce these polyketides is described and the critical steps of mobilizing large biosynthetic clusters and producing new linear cloning vectors are illustrated by experimental data.

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.

Enzymes of the shikimic acid pathway encoded in the genome of a basal metazoan, Nematostella vectensis, have microbial origins.

The shikimic acid pathway is responsible for the biosynthesis of many aromatic compounds by a broad range of organisms, including bacteria, fungi, plants, and some protozoans. Animals are considered to lack this pathway, as evinced by their dietary requirement for shikimate-derived aromatic amino acids. We challenge the universality of this traditional view in this report of genes encoding enzymes for the shikimate pathway in an animal, the starlet sea anemone Nematostella vectensis. Molecular evidence establishes horizontal transfer of ancestral genes of the shikimic acid pathway into the N. vectensis genome from both bacterial and eukaryotic (dinoflagellate) donors. Bioinformatic analysis also reveals four genes that are closely related to those of Tenacibaculum sp. MED152, raising speculation for the existence of a previously unsuspected bacterial symbiont. Indeed, the genome of the holobiont (i.e., the entity consisting of the host and its symbionts) comprises a high content of Tenacibaculum-like gene orthologs, including a 16S rRNA sequence that establishes the phylogenetic position of this associate to be within the family Flavobacteriaceae. These results provide a complementary view for the biogenesis of shikimate-related metabolites in marine Cnidaria as a "shared metabolic adaptation" between the partners.

Genetics of Streptomyces rimosus, the oxytetracycline producer.

From a genetic standpoint, Streptomyces rimosus is arguably the best-characterized industrial streptomycete as the producer of oxytetracycline and other tetracycline antibiotics. Although resistance to these antibiotics has reduced their clinical use in recent years, tetracyclines have an increasing role in the treatment of emerging infections and noninfective diseases. Procedures for in vivo and in vitro genetic manipulations in S. rimosus have been developed since the 1950s and applied to study the genetic instability of S. rimosus strains and for the molecular cloning and characterization of genes involved in oxytetracycline biosynthesis. Recent advances in the methodology of genome sequencing bring the realistic prospect of obtaining the genome sequence of S. rimosus in the near term.

Global genome analysis of the shikimic acid pathway reveals greater gene loss in host-associated than in free-living bacteria.

BACKGROUND: A central tenet in biochemistry for over 50 years has held that microorganisms, plants and, more recently, certain apicomplexan parasites synthesize essential aromatic compounds via elaboration of a complete shikimic acid pathway, whereas metazoans lacking this pathway require a dietary source of these compounds. The large number of sequenced bacterial and archaean genomes now available for comparative genomic analyses allows the fundamentals of this contention to be tested in prokaryotes. Using Hidden Markov Model profiles (HMM profiles) to identify all known enzymes of the pathway, we report the presence of genes encoding shikimate pathway enzymes in the hypothetical proteomes constructed from the genomes of 488 sequenced prokaryotes. RESULTS: Amongst free-living prokaryotes most Bacteria possess, as expected, genes encoding a complete shikimic acid pathway, whereas of the culturable Archaea, only one was found to have a complete complement of recognisable enzymes in its predicted proteome. It may be that in the Archaea, the primary amino-acid sequences of enzymes of the pathway are highly divergent and so are not detected by HMM profiles. Alternatively, structurally unrelated (non-orthologous) proteins might be performing the same biochemical functions as those encoding recognized genes of the shikimate pathway. Most surprisingly, 30% of host-associated (mutualistic, commensal and pathogenic) bacteria likewise do not possess a complete shikimic acid pathway. Many of these microbes show some degree of genome reduction, suggesting that these host-associated bacteria might sequester essential aromatic compounds from a parasitised host, as a 'shared metabolic adaptation' in mutualistic symbiosis, or obtain them from other consorts having the complete biosynthetic pathway. The HMM results gave 84% agreement when compared against data in the highly curated BioCyc reference database of genomes and metabolic pathways. CONCLUSIONS: These results challenge the conventional belief that the shikimic acid pathway is universal and essential in prokaryotes. The possibilities that non-orthologous enzymes catalyse reactions in this pathway (especially in the Archaea), or that there exist specific uptake mechanisms for the acquisition of shikimate intermediates or essential pathway products, warrant further examination to better understand the precise metabolic attributes of host-beneficial and pathogenic bacteria.

ClustScan: an integrated program package for the semi-automatic annotation of modular biosynthetic gene clusters and in silico prediction of novel chemical structures.

The program package 'ClustScan' (Cluster Scanner) is designed for rapid, semi-automatic, annotation of DNA sequences encoding modular biosynthetic enzymes including polyketide synthases (PKS), non-ribosomal peptide synthetases (NRPS) and hybrid (PKS/NRPS) enzymes. The program displays the predicted chemical structures of products as well as allowing export of the structures in a standard format for analyses with other programs. Recent advances in understanding of enzyme function are incorporated to make knowledge-based predictions about the stereochemistry of products. The program structure allows easy incorporation of additional knowledge about domain specificities and function. The results of analyses are presented to the user in a graphical interface, which also allows easy editing of the predictions to incorporate user experience. The versatility of this program package has been demonstrated by annotating biochemical pathways in microbial, invertebrate animal and metagenomic datasets. The speed and convenience of the package allows the annotation of all PKS and NRPS clusters in a complete Actinobacteria genome in 2-3 man hours. The open architecture of ClustScan allows easy integration with other programs, facilitating further analyses of results, which is useful for a broad range of researchers in the chemical and biological sciences.

Clustering of protein domains for functional and evolutionary studies.

BACKGROUND: The number of protein family members defined by DNA sequencing is usually much larger than those characterised experimentally. This paper describes a method to divide protein families into subtypes purely on sequence criteria. Comparison with experimental data allows an independent test of the quality of the clustering. RESULTS: An evolutionary split statistic is calculated for each column in a protein multiple sequence alignment; the statistic has a larger value when a column is better described by an evolutionary model that assumes clustering around two or more amino acids rather than a single amino acid. The user selects columns (typically the top ranked columns) to construct a motif. The motif is used to divide the family into subtypes using a stochastic optimization procedure related to the deterministic annealing EM algorithm (DAEM), which yields a specificity score showing how well each family member is assigned to a subtype. The clustering obtained is not strongly dependent on the number of amino acids chosen for the motif. The robustness of this method was demonstrated using six well characterized protein families: nucleotidyl cyclase, protein kinase, dehydrogenase, two polyketide synthase domains and small heat shock proteins. Phylogenetic trees did not allow accurate clustering for three of the six families. CONCLUSION: The method clustered the families into functional subtypes with an accuracy of 90 to 100%. False assignments usually had a low specificity score.

New methods for medicinal chemistry--universal gene cloning and expression systems for production of marine bioactive metabolites.

Natural products from symbiotic or commensal associations between marine invertebrate and microbial organisms show exceptional promise as pharmaceuticals in many therapeutic areas. An economic and sustainable global market supply due to difficulty of synthesis is cited as the main obstacle for exploitation of these otherwise exciting marine bioactive compounds. Different strategies have been evoked to overcome this impediment as long-term harvesting of wild stocks from the environment is considered unsound, and other modes of production based on biosynthesis, such as aquaculture, have not yet been proven as reliable. One option is to clone the genes encoding the biosynthetic expression of a lead metabolite into a surrogate host suitable for industrial-scale fermentation. To facilitate this goal we are developing a universal system to clone and express genes responsible for biosynthesis of natural products from both eukaryotic and prokaryotic partners of marine symbioses. The ability to harness the complete meta-transcriptome of entire biosynthetic pathways is particularly valuable where the biogenesis of a target natural product occurring within a complex symbiotic association is unclear.

Plasticity of the streptomyces genome-evolution and engineering of new antibiotics.

Streptomyces is a genus of soil dwelling bacteria with the ability to produce natural products that have found widespread use in medicine. Annotation of Streptomyces genome sequences has revealed far more biosynthetic gene clusters than previously imagined, offering exciting possibilities for future combinatorial biosynthesis. Experiments to manipulate modular biosynthetic clusters to create novel chemistries often result in no detectable product or product yield is extremely low. Understanding the coupling between components in these hybrid enzymes will be crucial for efficient synthesis of new compounds. We are using new algebraic approaches to predict protein properties, and homologous recombination to exploit natural evolutionary constraints to generate novel functional enzymes. The methods and techniques developed could easily be adapted to study modular, multi-interacting complex systems where appreciable biochemical and comparative sequence data are available, for example, clinically significant non-ribosomally synthesised peptides and polyketides.