Role of linkers in communication between protein modules

Improved general regression network for protein domain boundary prediction

Background

Protein domains present some of the most useful information that can be used to understand protein structure and functions. Recent research on protein domain boundary prediction has been mainly based on widely known machine learning techniques, such as Artificial Neural Networks and Support Vector Machines. In this study, we propose a new machine learning model (IGRN) that can achieve accurate and reliable classification, with significantly reduced computations. The IGRN was trained using a PSSM (Position Specific Scoring Matrix), secondary structure, solvent accessibility information and inter-domain linker index to detect possible domain boundaries for a target sequence.

Results

The proposed model achieved average prediction accuracy of 67% on the Benchmark_2 dataset for domain boundary identification in multi-domains proteins and showed superior predictive performance and generalisation ability among the most widely used neural network models. With the CASP7 benchmark dataset, it also demonstrated comparable performance to existing domain boundary predictors such as DOMpro, DomPred, DomSSEA, DomCut and DomainDiscovery with 70.10% prediction accuracy.

Conclusion

The performance of proposed model has been compared favourably to the performance of other existing machine learning based methods as well as widely known domain boundary predictors on two benchmark datasets and excels in the identification of domain boundaries in terms of model bias, generalisation and computational requirements.

The linker region plays a key role in the adaptation to cold of the cellulase from an Antarctic bacterium

The psychrophilic cellulase, Cel5G, from the Antarctic bacterium Pseudoalteromonas haloplanktis is composed of a catalytic module (CM) joined to a carbohydrate-binding module (CBM) by an unusually long, extended and flexible linker region (LR) containing three loops closed by three disulfide bridges. To evaluate the possible role of this region in cold adaptation, the LR was sequentially shortened by protein engineering, successively deleting one and two loops of this module, whereas the last disulfide bridge was also suppressed by replacing the last two cysteine residue by two alanine residues. The kinetic and thermodynamic properties of the mutants were compared with those of the full-length enzyme, and also with those of the cold-adapted CM alone and with those of the homologous mesophilic enzyme, Cel5A, from Erwinia chrysanthemi. The thermostability of the mutated enzymes as well as their relative flexibility were evaluated by differential scanning calorimetry and fluorescence quenching respectively. The topology of the structure of the shortest mutant was determined by SAXS (small-angle X-ray scattering). The data indicate that the sequential shortening of the LR induces a regular decrease of the specific activity towards macromolecular substrates, reduces the relative flexibility and concomitantly increases the thermostability of the shortened enzymes. This demonstrates that the long LR of the full-length enzyme favours the catalytic efficiency at low and moderate temperatures by rendering the structure not only less compact, but also less stable, and plays a crucial role in the adaptation to cold of this cellulolytic enzyme.

Folate-mediated one-carbon metabolism

Tetrahydrofolate (THF) polyglutamates are a family of cofactors that carry and chemically activate one-carbon units for biosynthesis. THF-mediated one-carbon metabolism is a metabolic network of interdependent biosynthetic pathways that is compartmentalized in the cytoplasm, mitochondria, and nucleus. One-carbon metabolism in the cytoplasm is required for the synthesis of purines and thymidylate and the remethylation of homocysteine to methionine. One-carbon metabolism in the mitochondria is required for the synthesis of formylated methionyl-tRNA; the catabolism of choline, purines, and histidine; and the interconversion of serine and glycine. Mitochondria are also the primary source of one-carbon units for cytoplasmic metabolism. Increasing evidence indicates that folate-dependent de novo thymidylate biosynthesis occurs in the nucleus of certain cell types. Disruption of folate-mediated one-carbon metabolism is associated with many pathologies and developmental anomalies, yet the biochemical mechanisms and causal metabolic pathways responsible for the initiation and/or progression of folate-associated pathologies have yet to be established. This chapter focuses on our current understanding of mammalian folate-mediated one-carbon metabolism, its cellular compartmentation, and knowledge gaps that limit our understanding of one-carbon metabolism and its regulation.

Multienzyme docking in hybrid megasynthetases

Hybrid multienzyme systems composed of polyketide synthase (PKS) and nonribosomal polypeptide synthetase (NRPS) modules direct the biosynthesis of clinically valuable natural products in bacteria. The fidelity of this process depends on specific recognition between successive polypeptides in each assembly line-interactions that are mediated by terminal 'docking domains'. We have identified a new family of N-terminal docking domains, exemplified by TubCdd from the tubulysin system of Angiococcus disciformis An d48. TubCdd is homodimeric, which suggests that NRPS subunits in mixed systems self-associate to interact with partner PKS homodimers. The NMR structure of TubCdd reveals a new fold featuring an exposed beta-hairpin that serves as the binding site for the C-terminal docking domain of the partner polypeptide. The pattern of charged residues on the contact surface of the beta-hairpin is a key determinant of the interaction and seems to constitute a 'docking code' that can be used to alter binding affinity.

Versatility of polyketide synthases in generating metabolic diversity

Polyketide synthases (PKSs) form a large family of multifunctional proteins involved in the biosynthesis of diverse classes of natural products. Architecturally at least three different types of PKSs have been discovered in the microbial world and recent years have revealed tremendous versatility of PKSs, both in terms of their structural and functional organization and in their ability to produce compounds other than typical secondary metabolites. Mycobacterium tuberculosis exploits polyketide biosynthetic enzymes to synthesize complex lipids, many of which are essential for its survival. The functional significance of the large repertoire of PKSs in Dictyostelium discoideum, perhaps in producing developmental regulating factors, is emerging. Recently determined structures of fatty acid synthases (FASs) and PKSs now provide an opportunity to delineate the mechanistic and structural basis of polyketide biosynthetic machinery.

Interrogating the molecular basis for multiple macrolactone ring formation by the pikromycin polyketide synthase

The pikromycin polyketide synthase (PKS) is unique in its ability to generate both 12 and 14 membered ring macrolactones. As such, dissection of the molecular basis for controlling metabolic diversity in this system remains an important objective for understanding modular PKS function and expanding chemical diversity. Here, we describe a series of experiments designed to probe the importance of the protein-protein interaction that occurs between the final two monomodules, PikAIII (module 5) and PikAIV (module 6), for the production of the 12 membered ring macrolactone 10-deoxymethynolide. The results obtained from these in vitro studies demonstrate that PikAIII and PikAIV generate the 12 membered ring macrocycle most efficiently when engaged in their native protein-protein interaction. Accordingly, the data are consistent with PikAIV adopting an alternative conformation that enables the terminal thioesterase domain to directly off-load the PikAIII-bound hexaketide intermediate for macrocyclization.

Versatile polyketide enzymatic machinery for the biosynthesis of complex mycobacterial lipids

The cell envelope of Mycobacterium tuberculosis (Mtb) is a treasure house of a variety of biologically active molecules with fascinating architectures. The decoding of the genetic blueprint of Mtb in recent years has provided the impetus for dissecting the metabolic pathways involved in the biosynthesis of lipidic metabolites. The focus of the Highlight is to emphasize the functional role of polyketide synthase (PKS) proteins in the biosynthesis of complex mycobacterial lipids. The catalytic as well as mechanistic versatility of PKS. in generating metabolic diversity and the significance of recently discovered fatty acyl-AMP ligases in establishing "biochemical crosstalk" between fatty acid synthases (FASs) and PKSs is described. The phenotypic heterogeneity and remodeling of the mycobacterial cell wall in its aetiopathogenesis is discussed.

Dephosphorylation of phosphotyrosine on STAT1 dimers requires extensive spatial reorientation of the monomers facilitated by the N-terminal domain

We report experiments that infer a radical reorientation of tyrosine-phosphorylated parallel STAT1 dimers to an antiparallel form. Such a change in structure allows easy access to a phosphatase. With differentially epitope-tagged molecules, we show that the two monomers of a dimer remain together during dephosphorylation although they most likely undergo spatial reorientation. Extensive single amino acid mutagenesis within crystallographically established domains, manipulation of amino acids in an unstructured tether that connects the N-terminal domain (ND) to the core of the protein, and the demonstration that overexpressed ND can facilitate dephosphorylation of a core molecule lacking an ND all support this model: When the tyrosine-phosphorylated STAT1 disengages from DNA, the ND dimerizes and somehow assists in freeing the reciprocal pY-SH2 binding between the monomers of the dimer while ND ND dimerization persists. The core of the monomers rotate allowing reciprocal association of the coiled:coil and DNA-binding domains to present pY at the two ends of an antiparallel dimer for ready dephosphorylation.

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.

In Vivo Biocombinatorial Synthesis of Lipopeptides by COM Domain-Mediated Reprogramming of the Surfactin Biosynthetic Complex

The intermolecular communication within NRPS complexes relies on the coordinated interplay of donor and acceptor communication-mediating (COM) domains. In this study, the potential of COM domains was exploited in vivo by establishing a system for the true biocombinatorial synthesis of lipopeptides via directed reprogramming of a natural NRP biosynthetic assembly line (i.e., surfactin). By means of COM domain swapping, these experiments verified the decisive role of COM domains for the control of protein-protein interactions between NRPSs, demonstrated the functionality of COM domain pairs even in the context of a heterologous host and NRPS system, and allowed for the intended skipping of a biosynthetic enzyme within a multienzymatic biosynthetic complex. Ultimately, abrogation of the selectivity barrier provided by COM domains afforded the successful simultaneous, biocombinatorial synthesis of distinct lipopeptide products.

Amplification of DNA encoding entire type I polyketide synthase domains and linkers from streptomyces species

Polyketides are a group of bioactive compounds from bacteria, plants, and fungi. To increase the availability of analogs for testing, the active sites of polyketide synthases are often substituted with homologous domains having altered substrate specificities. This study reports the design of polymerase chain reaction primers that enables isolation of entire active site domains from type I polyketide synthases with native interdomain linkers. This bypasses the need for further genetic screening to obtain functional units for use in genetic engineering. This is especially important in bioprospecting projects exploring new environments for bioresources.

Control of protein functional dynamics by peptide linkers

Control of structural flexibility is essential for the proper functioning of a large number of proteins and multiprotein complexes. At the residue level, such flexibility occurs due to local relaxation of peptide bond angles whose cumulative effect may result in large changes in the secondary, tertiary or quaternary structures of protein molecules. Such flexibility, and its absence, most often depends on the nature of interdomain linkages formed by oligopeptides. Both flexible and relatively rigid peptide linkers are found in many multidomain proteins. Linkers are thought to control favorable and unfavorable interactions between adjacent domains by means of variable softness furnished by their primary sequence. Large-scale structural heterogeneity of multidomain proteins and their complexes, facilitated by soft peptide linkers, is now seen as the norm rather than the exception. Biophysical discoveries as well as computational algorithms and databases have reshaped our understanding of the often spectacular biomolecular dynamics enabled by soft linkers. Absence of such motion, as in so-called molecular rulers, also has desirable functional effects in protein architecture. We review here the historic discovery and current understanding of the nature of domains and their linkers from a structural, computational, and biophysical point of view. A number of emerging applications, based on the current understanding of the structural properties of peptides, are presented in the context of domain fusion of synthetic multifunctional chimeric proteins.

The Structural Basis for Docking in Modular Polyketide Biosynthesis

Polyketide natural products such as erythromycin and rapamycin are assembled on polyketide synthases (PKSs), which consist of modular sets of catalytic activities distributed across multiple protein subunits. Correct protein-protein interactions among the PKS subunits which are critical to the fidelity of biosynthesis are mediated in part by "docking domains" at the termini of the proteins. The NMR solution structure of a representative docking domain complex from the erythromycin PKS (DEBS) was recently solved, and on this basis it has been proposed that PKS docking is mediated by the formation of an intermolecular four-alpha-helix bundle. Herein, we report the genetic engineering of such a docking domain complex by replacement of specific helical segments and analysis of triketide synthesis by mutant PKSs in vivo. The results of these helix swaps are fully consistent with the model and highlight residues in the docking domains that may be targeted to alter the efficiency or specificity of subunit-subunit docking in hybrid PKSs.

The expression of selected non-ribosomal peptide synthetases in Aspergillus fumigatus is controlled by the availability of free iron

Three non-ribosomal peptide synthetase genes, termed sidD, sidC and sidE, have been identified in Aspergillus fumigatus. Gene expression analysis by RT-PCR confirms that expression of both sidD and C was reduced by up to 90% under iron-replete conditions indicative of a likely role in siderophore biosynthesis. SidE expression was less sensitive to iron levels. In addition, two proteins purified from mycelia grown under iron-limiting conditions corresponded to SidD ( approximately 200 kDa) and SidC (496 kDa) as determined by MALDI ToF peptide mass fingerprinting and MALDI LIFT-ToF/ToF. Siderophore synthetases are unique in bacteria and fungi and represent an attractive target for antimicrobial chemotherapy.

Armadillo: domain boundary prediction by amino acid composition

The identification and annotation of protein domains provides a critical step in the accurate determination of molecular function. Both computational and experimental methods of protein structure determination may be deterred by large multi-domain proteins or flexible linker regions. Knowledge of domains and their boundaries may reduce the experimental cost of protein structure determination by allowing researchers to work on a set of smaller and possibly more successful alternatives. Current domain prediction methods often rely on sequence similarity to conserved domains and as such are poorly suited to detect domain structure in poorly conserved or orphan proteins. We present here a simple computational method to identify protein domain linkers and their boundaries from sequence information alone. Our domain predictor, Armadillo (http://armadillo.blueprint.org), uses any amino acid index to convert a protein sequence to a smoothed numeric profile from which domains and domain boundaries may be predicted. We derived an amino acid index called the domain linker propensity index (DLI) from the amino acid composition of domain linkers using a non-redundant structure dataset. The index indicates that Pro and Gly show a propensity for linker residues while small hydrophobic residues do not. Armadillo predicts domain linker boundaries from Z-score distributions and obtains 35% sensitivity with DLI in a two-domain, single-linker dataset (within +/-20 residues from linker). The combination of DLI and an entropy-based amino acid index increases the overall Armadillo sensitivity to 56% for two domain proteins. Moreover, Armadillo achieves 37% sensitivity for multi-domain proteins, surpassing most other prediction methods. Armadillo provides a simple, but effective method by which prediction of domain boundaries can be obtained with reasonable sensitivity. Armadillo should prove to be a valuable tool for rapidly delineating protein domains in poorly conserved proteins or those with no sequence neighbors. As a first-line predictor, domain meta-predictors could yield improved results with Armadillo predictions.

Type I polyketide synthases may have evolved through horizontal gene transfer

Type I polyketide synthases (PKSI) are modular multidomain enzymes involved in the biosynthesis of many natural products of industrial interest. PKSI modules are minimally organized in three domains: ketosynthase (KS), acyltransferase (AT), and acyl carrier protein. The KS domain phylogeny of 23 PKSI clusters was determined. The results obtained suggest that many horizontal transfers of PKSI genes have occurred between actinomycetales species. Such gene transfers may explain the homogeneity and the robustness of the actinomycetales group since gene transfers between closely related species could mimic patterns generated by vertical inheritance. We suggest that the linearity and instability of actinomycetales chromosomes associated with their large quantity of genetic mobile elements have favored such horizontal gene transfers.

Prediction of protein interdomain linker regions by a hidden Markov model

MOTIVATION

Our aim was to predict protein interdomain linker regions using sequence alone, without requiring known homology. Identifying linker regions will delineate domain boundaries, and can be used to computationally dissect proteins into domains prior to clustering them into families. We developed a hidden Markov model of linker/non-linker sequence regions using a linker index derived from amino acid propensity. We employed an efficient Bayesian estimation of the model using Markov Chain Monte Carlo, Gibbs sampling in particular, to simulate parameters from the posteriors. Our model recognizes sequence data to be continuous rather than categorical, and generates a probabilistic output.

RESULTS

We applied our method to a dataset of protein sequences in which domains and interdomain linkers had been delineated using the Pfam-A database. The prediction results are superior to a simpler method that also uses linker index.

Preparation and characterization of a single-chain calcineurin–calmodulin complex

Calcineurin (CN), a Ca(2+)/calmodulin (CaM)-dependent serine/threonine protein phosphatase, is a heterodimer composed of a catalytic subunit (CNA) and a regulatory subunit (CNB). The activity of CNA is under the control of two functionally distinct, but structurally similar Ca(2+)-regulated proteins, CaM and CNB. The crystal structure of the holoenzyme reveals that the N-terminus and C-terminus of CNB and the N-terminus of CNA each have a long arm not involved in the active site. We constructed a fusion of the genes of CaM, CNB and CNA in that order using linker primers containing six and ten codons of glycine. A single-chain CaM-CNB-CNA (CBA) complex was expressed and purified to near homogeneity. The single-chain complex was fully soluble, and had biochemical properties and kinetic parameters similar to single-chain CNB-CNA (BA) activated by CaM. It was not regulated by CaM and CNB, but was strongly stimulated by Mn2+, Ni2+ and Mg2+. Intrinsic fluorescence spectroscopy of the complex showed a change in the environment of tryptophan in the presence of Ca2+ and circular dichroism (CD) spectropolarimetry revealed an increase in alpha-helical content. Our findings suggest that fusion of CaM, CNB and CNA does not prevent the structural changes required for their functioning; in particular, CaM within the complex could still interact correctly with CN in the presence of Ca2+.

Polyketide biosynthesis: understanding and exploiting modularity

Polyketide-based pharmaceuticals are some of our most important medicines. They are constructed in micro-organisms (typically bacteria and fungi) by gigantic enzyme catalysts called polyketide synthases (PKSs). The organization of PKSs into molecular assembly lines makes them particularly appealing targets for genetic engineering because, in principle, an alteration in the enzyme organization might translate into a predictable change in polyketide structure. Excitingly, this has been shown repeatedly to work in practice, but the efficiency of the engineered PKSs is frequently too low to be useful for large-scale drug synthesis. To reach this goal, researchers need a deeper understanding of the structure and function of these proteins, which are among the most complex in nature. This review highlights some recent experiments which are providing key information about the molecular organization, mechanism and orchestration of these magnificent catalysts, and opening up fresh prospects of truly combinatorial biosynthesis of novel polyketides as leads in drug discovery.

Hybrid Nonribosomal Peptide-Polyketide Interfaces in Epothilone Biosynthesis

Epothilone (Epo) D, an antitumor agent currently in clinical trials, is a hybrid natural product produced by the combined action of nonribosomal peptide synthetases (NRPS) and polyketide synthases (PKS). In the epothilone biosynthetic pathway, EpoB, a 165 kDa NRPS is inserted into an otherwise entirely PKS assembly line, forming two hybrid NRPS-PKS interfaces. In light of the terminal linker effect previously identified in PKS, the N- and C-terminal sequences of EpoB were examined for their roles in propagating the incipient natural product. Eight amino acid residues at EpoB C terminus, in which six are positively charged, were found to be a key component of the C-terminal linker effect. A minimal sequence of 56 residues at EpoB N terminus was required for elongating the acetyl group from the acyl carrier protein (ACP) of EpoA to form methylthiazolyl-S-EpoB.

Don't classify polyketide synthases

Polyketide synthases are intensively studied as metabolite factories generating diverse biologically active natural products. Contrary to their current classification as different "types," there is now a growing body of evidence illustrating that nature realized limitless transitional stages during evolution.

NRPS-PKS: a knowledge-based resource for analysis of NRPS/PKS megasynthases

NRPS-PKS is web-based software for analysing large multi-enzymatic, multi-domain megasynthases that are involved in the biosynthesis of pharmaceutically important natural products such as cyclosporin, rifamycin and erythromycin. NRPS-PKS has been developed based on a comprehensive analysis of the sequence and structural features of several experimentally characterized biosynthetic gene clusters. The results of these analyses have been organized as four integrated searchable databases for elucidating domain organization and substrate specificity of nonribosomal peptide synthetases and three types of polyketide synthases. These databases work as the backend of NRPS-PKS and provide the knowledge base for predicting domain organization and substrate specificity of uncharacterized NRPS/PKS clusters. Benchmarking on a large set of biosynthetic gene clusters has demonstrated that, apart from correct identification of NRPS and PKS domains, NRPS-PKS can also predict specificities of adenylation and acyltransferase domains with reasonably high accuracy. These features of NRPS-PKS make it a valuable resource for identification of natural products biosynthesized by NRPS/PKS gene clusters found in newly sequenced genomes. The training and test sets of gene clusters included in NRPS-PKS correlate information on 307 open reading frames, 2223 functional protein domains, 68 starter/extender precursors and their specific recognition motifs, and also the chemical structure of 101 natural products from four different families. NRPS-PKS is a unique resource which provides a user-friendly interface for correlating chemical structures of natural products with the domains and modules in the corresponding nonribosomal peptide synthetases or polyketide synthases. It also provides guidelines for domain/module swapping as well as site-directed mutagenesis experiments to engineer biosynthesis of novel natural products. NRPS-PKS can be accessed at http://www.nii.res.in/nrps-pks.html.

Macrotetrolide biosynthesis: a novel type II polyketide synthase

Polyketide biosynthesis is catalyzed by polyketide synthase (PKS) and three types of bacterial PKS are known to date. Feeding experiments with isotope-labeled precursors established the polyketide origin of the macrotetrolides, but the labeling pattern cannot be rationalized according to the established PKS paradigm. Genetic analysis of the macrotetrolide biosynthesis unveiled an unprecedented organization for a polyketide gene cluster that features five genes encoding discrete ketoacyl synthase (KS) and four genes encoding discrete ketoreductase (KR) but lacking an acyl carrier protein (ACP). Macrotetrolide biosynthesis is proposed to involve a novel type II PKS that acts directly on acyl CoA substrates, functions noniteratively, and catalyzes both C-C and C-O bond formation. These findings demonstrate once again Nature's versatility in making complex molecules and suggests new strategies for PKS engineering to further expand the scope and diversity of polyketide library. They also should serve as an inspiration in searching for PKS with novel chemistry for combinatorial biosynthesis.

Engineering a bifunctional starch–cellulose cross-bridge protein

Biodegradable starch- and cellulose-based polymers have a range of properties which make them suitable for use in a wide array of biomedical applications ranging from bone replacement to engineering of tissue scaffolds and drug delivery systems. A novel polysaccharide cross-bridging protein was designed which was comprised of a cellulose-binding domain from Clostridium cellulovorans (CBD(clos)) and a starch-binding domain from Aspergillus niger B1 (SBD(Asp)). The two genes were fused in-frame via a synthetic elastin gene to construct a Cellulose/Starch Cross bridging Protein (CSCP). Recombinant CSCP was expressed in Escherichia coli, and successfully refolded from inclusion bodies. CSCP demonstrated cross-bridging ability in different model systems composed of insoluble or soluble starch and cellulose. The aspect that different carbohydrate-binding module maintain their binding capacity over a wide range of conditions, without the need for chemical reactions, makes them attractive domains for designing new classes of chimeric polysaccharide-binding domains which demonstrate potential for use in a wide range of biomaterials.

Structures of the N-terminal modules imply large domain motions during catalysis by methionine synthase

B(12)-dependent methionine synthase (MetH) is a large modular enzyme that utilizes the cobalamin cofactor as a methyl donor or acceptor in three separate reactions. Each methyl transfer occurs at a different substrate-binding domain and requires a different arrangement of modules. In the catalytic cycle, the cobalamin-binding domain carries methylcobalamin to the homocysteine (Hcy) domain to form methionine and returns cob(I)alamin to the folate (Fol) domain for remethylation by methyltetrahydrofolate (CH(3)-H(4)folate). Here, we describe crystal structures of a fragment of MetH from Thermotoga maritima comprising the domains that bind Hcy and CH(3)-H(4)folate. These substrate-binding domains are (beta alpha)(8) barrels packed tightly against one another with their barrel axes perpendicular. The properties of the domain interface suggest that the two barrels remain associated during catalysis. The Hcy and CH(3)-H(4)folate substrates are bound at the C termini of their respective barrels in orientations that position them for reaction with cobalamin, but the two active sites are separated by approximately 50 A. To complete the catalytic cycle, the cobalamin-binding domain must travel back and forth between these distant active sites.

The structure of docking domains in modular polyketide synthases

Polyketides from actinomycete bacteria provide the basis for many valuable medicines, so engineering genes for their biosynthesis to produce variant molecules holds promise for drug discovery. The modular polyketide synthases are particularly amenable to this approach, because each cycle of chain extension is catalyzed by a different module of enzymes, and the modules are arranged within giant multienzyme subunits in the order in which they act. Protein-protein interactions between terminal docking domains of successive multienzymes promote their correct positioning within the assembly line, but because the overall complex is not stable in vitro, the key interactions have not been identified. We present here the NMR solution structure of a 120 residue polypeptide representing a typical pair of such domains, fused at their respective C and N termini: it adopts a stable dimeric structure which reveals the detailed role of these (predominantly helical) domains in docking and dimerization by modular polyketide synthases.

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.

Identification and cloning of genes encoding viomycin biosynthesis from Streptomyces vinaceus and evidence for involvement of a rare oxygenase

The tuberactinomycins are a family of basic cyclic peptides that exhibit potent antitubercular activity. These peptides are characterized by the presence of an amino acid with a 6-membered cyclic guanidine side chain (capreomycidine) and two or more 2,3-diaminopropionate residues. Viomycin (tuberactinomycin B) is a well-studied member of the family, was once prescribed for the treatment of tuberculosis, and has been shown to block translocation during protein biosynthesis. The gene cluster encoding viomycin biosynthesis was identified and cloned from Streptomyces vinaceus. The cluster was identified by screening genomic libraries with the viomycin phosphotransferase self-resistance gene (vph) and non-ribosomal peptide synthetase (NRPS) gene probes amplified from S. vinaceus genomic DNA. The viomycin cluster was localized to ca. 120 kb of contiguous DNA defined by four overlapping cosmid inserts. Each cosmid hybridized with one or more peptide synthetase gene probes and two also hybridized with vph. Confirmation that the cluster encoded viomycin biosynthesis was obtained from the disruption of two NRPS adenylation domains. Partial sequence analysis revealed an ORF (svox) predicted to encode a rare non-heme iron, alpha-ketoglutarate dependent oxygenase proposed to function in the oxidative cyclization of arginine to the capreomycidine residue. Insertional disruption of svox resulted in complete loss of viomycin production, confirming its involvement in the pathway.

Inhibition of HBV targeted ribonuclease enhanced by introduction of linker

AIM: To construct human eosinophil-derived neurotoxin(hEDN) and HBV core protein (HBVc) eukaryotic fusion expression vector with a linker (Gly₄Ser)₃ between them to optimize the molecule folding, which will be used to inhibit HBV replication in vitro.

METHODS: Previously constructed pcDNA3.1(-)/TR was used as a template. Linker sequence was synthesized and annealed to form dslinker, and cloned into pcDNA3.1(-)/TR to produce plasmid pcDNA3.1(-)/HBc-linker. Then the hEDN fragment was PCR amplified and inserted into pcDNA3.1(-)/HBc-linker to form pcDNA3.1(-)/TNL in which the effector molecule and the target molecule were separated by a linker sequence. pcDNA3.1(-)/TNL expression was identified by indirect immunofluorescence staining. Radioimmunoassay was used to analyse anti-HBV activity of pcDNA3.1(-)/TNL. Meanwhile, metabolism of cells was evaluated by MTT colorimetry.

RESULTS: hEDN and HBVc eukaryotic fusion expression vector with a linker (Gly4Ser)3 between them was successfully constructed. pcDNA3.1(-)/TNL was expressed in HepG2.2.15 cells efficiently. A significant decrease of HBsAg concentration from pcDNA3.1(-)/TNL transfectant was observed compared to pcDNA3.1(-)/TR (P = 0.036, P < 0.05). MTT assay suggested that there were no significant differences between groups (P = 0.08, P > 0.05).

CONCLUSION: Linker introduction enhances the inhibitory effect of HBV targeted ribonuclease significantly.

SEARCHPKS: A program for detection and analysis of polyketide synthase domains

SEARCHPKS is a software for detection and analysis of polyketide synthase (PKS) domains in a polypeptide sequence. Modular polyketide synthases are unusually large multi-enzymatic multi-domain megasynthases, which are involved in the biosynthesis of pharmaceutically important natural products using an assembly-line mechanism. This program facilitates easy identification of various PKS domains and modules from a given polypeptide sequence. In addition, it also predicts the specificity of the potential acyltransferase domains for various starter and extender precursor units. SEARCHPKS is a user-friendly tool for correlating polyketide chemical structures with the organization of domains and modules in the corresponding modular polyketide synthases. This program also allows the user to extensively analyze and assess the sequence homology of various polyketide synthase domains, thus providing guidelines for carrying out domain and module swapping experiments. SEARCHPKS can also aid in identification of polyketide products made by PKS clusters found in newly sequenced genomes. The computational approach used in SEARCHPKS is based on a comprehensive analysis of various characterized clusters of modular polyketide synthases compiled in PKSDB, a database of modular polyketide synthases. SEARCHPKS can be accessed at http://www.nii.res.in/searchpks.html.

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.

Photoaffinity labeling of the N-methyltransferase domains of cyclosporin synthetase

The multifunctional polypeptide cyclosporin synthetase (CySyn) remains one of the most complex nonribosomal peptide synthetase described. In this study we used a highly specific photoaffinity labeling procedure with the natural cofactor S-adenosyl-L-methionine (AdoMet), 14C-isotopically labeled at the Sdelta methyl group to probe the concerted AdoMet-binding interaction of the N-methyltransferase (N-MTase) centers of CySyn. The binding stoichiometry for the enzyme-AdoMet complex was determined to be 1:7, which is in agreement with inferences made from analysis of the complementary DNA sequence of the simA gene encoding the CySyn polypeptide. The photolabeling of the AdoMet-binding sites displayed homotropic negative cooperativity, characterized by a curvilinear Scatchard plot with upward concavity. Although, the process of N-methyl transfer is not a critical event for peptide elongation, the destabilizing homotropic interactions between N-MTase centers that were observed may represent a mechanism whereby the enzyme preserves the proficiency of the substrate-channeling process of cyclosporin peptide assembly over a broad range of cofactor concentrations. Furthermore, we demonstrated the utility of the photolabeling procedure for tracking the enzyme during purification.

Mapping and molecular modeling of S-adenosyl-L-methionine binding sites in N-methyltransferase domains of the multifunctional polypeptide cyclosporin synthetase

We employed a highly specific photoaffinity labeling procedure, using (14)C-labeled S-adenosyl-l-methionine (AdoMet) to define the chemical structure of the AdoMet binding centers on cyclosporin synthetase (CySyn). Tryptic digestion of CySyn photolabeled with either [methyl-(14)C]AdoMet or [carboxyl-(14)C]AdoMet yielded the sequence H(2)N-Asn-Asp-Gly-Leu-Glu-Ser-Tyr-Val-Gly-Ile-Glu-Pro-Ser-Arg-COOH (residues 10644-10657), situated within the N-methyltransferase domain of module 8 of CySyn. Radiosequencing detected Glu(10654) and Pro(10655) as the major sites of derivatization. [carboxyl-(14)C]AdoMet in addition labeled Tyr(10650). Chymotryptic digestion generated the radiolabeled peptide H(2)N-Ile-Gly-Leu-Glu-Pro-Ser-Gln-Ser-Ala-Val-Gln-Phe-COOH, corresponding to amino acids 2125-2136 of the N-methyltransferase domain of module 2. The radiolabeled amino acids were identified as Glu(2128) and Pro(2129), which are equivalent in position and function to the modified residues identified with tryptic digestions in module 8. Homology modeling of the N-methyltransferase domains indicates that these regions conserve the consensus topology of the AdoMet binding fold and consensus cofactor interactions seen in structurally characterized AdoMet-dependent methyltransferases. The modified sequence regions correspond to the motif II consensus sequence element, which is involved in directly complexing the adenine and ribose components of AdoMet. We conclude that the AdoMet binding to nonribosomal peptide synthetase N-methyltransferase domains obeys the consensus cofactor interactions seen among most structurally characterized low molecular weight AdoMet-dependent methyltransferases.

Evidence from engineered gene fusions for the repeated use of a module in a modular polyketide synthase

Functional evidence for programmed loss of co-linearity on the borrelidin modular polyketide synthase (PKS) is presented.

Modification of biologically active peptides: production of a novel lipohexapeptide after engineering of Bacillus subtilis surfactin synthetase

The Bacillus subtilis strain ATCC 21332 produces the lipoheptapeptide surfactin, a highly potent biosurfactant synthesized by a large multimodular peptide synthetase. We report the genetic engineering of the surfactin biosynthesis resulting in the production of a novel lipohexapeptide with altered antimicrobial activities. A combination of in vitro and in vivo recombination approaches was used to construct a modified peptide synthetase by eliminating a large internal region of the enzyme containing a complete amino acid incorporating module. The remaining modules adjacent to the deletion were recombined at different highly conserved sequence motifs characteristic of amino acid incorporating modules of peptide synthetases. The primary goal of this work was to identify permissive fusion sites suitable for the engineering of peptide synthetase genes by genetic recombination. Analysis of the rearranged enzymes after purification from B. subtilis and from the heterologous host Escherichia coli revealed that the selection of the recombination site is of crucial importance for a successful engineering. Only the recombination at a specific HHII x DGVS sequence motif resulted in an active peptide synthetase. The expected lipohexapeptide was produced in vivo and first evidence of a reduced toxicity against erythrocytes and an enhanced lysis of Bacillus licheniformis cells was shown.

An analysis of protein domain linkers: their classification and role in protein folding

Recent advances in protein engineering have come from creating multi-functional chimeric proteins containing modules from various proteins. These modules are typically joined via an oligopeptide linker, the correct design of which is crucial for the desired function of the chimeric protein. Here we analyse the properties of naturally occurring inter-domain linkers with the aim to design linkers for domain fusion. Two main types of linker were identified; helical and non-helical. Helical linkers are thought to act as rigid spacers separating two domains. Non-helical linkers are rich in prolines, which also leads to structural rigidity and isolation of the linker from the attached domains. This means that both linker types are likely to act as a scaffold to prevent unfavourable interactions between folding domains. Based on these results we have constructed a linker database intended for the rational design of linkers for domain fusion, which can be accessed via the Internet at http://mathbio.nimr.mrc.ac.uk.

The biosynthesis of the aromatic myxobacterial electron transport inhibitor stigmatellin is directed by a novel type of modular polyketide synthase

Deductions from the molecular analysis of the 65,000-bp stigmatellin biosynthetic gene cluster are reported. The biosynthetic genes (stiA-J) encode an unusual bacterial modular type I polyketide synthase (PKS) responsible for the formation of this aromatic electron transport inhibitor produced by the myxobacterium Stigmatella aurantiaca. Involvement of the PKS gene cluster in stigmatellin biosynthesis is shown using site-directed mutagenesis. One module of the PKS is assumed to be used iteratively during the biosynthetic process, which seems to involve an unusual transacylation of the biosynthetic intermediate from an acyl carrier protein domain back to the preceding ketosynthase domain. Finally, the polyketide chain which is presumably catalyzed by a novel C-terminal domain in StiJ that does not resemble thioesterases, is cyclized and aromatized. The presented results of feeding experiments are in good agreement with the proposed biosynthetic scheme. In contrast to all other PKS type I systems reported to date, each module of StiA-J is encoded on a separate gene. The gene cluster contains a "stand alone" O-methyltransferase and two unusual O-methyltransferase domains embedded in the PKS. In addition, inactivation of a cytochrome P450 monooxygenase-encoding gene involved in post-PKS hydroxylation of the aromatic ring leads to the formation of two novel stigmatellin derivatives.

Cloning and characterization of the bleomycin biosynthetic gene cluster from Streptomyces verticillus ATCC15003

Bleomycin (BLM) biosynthesis has been studied as a model for hybrid peptide-polyketide natural product biosynthesis. Cloning, sequencing, and biochemical characterization of the blm biosynthetic gene cluster from Streptomyces verticillus ATCC15003 revealed that (1) the BLM hybrid peptide-polyketide aglycon is assembled by the BLM megasynthetase that consists of both nonribosomal peptide synthetase (NRPS) and polyketide synthase (PKS) modules; (2) BlmIX/BlmVIII/BlmVII constitute a natural hybrid NRPS/PKS/NRPS system, serving as a model for both hybrid NRPS/PKS and PKS/NRPS systems; (3) the catalytic sites appear to be conserved in both hybrid NRPS/PKS and nonhybrid NRPS or PKS systems, with the exception of the KS domains in the hybrid NRPS/PKS systems that are unique; (4) specific interpolypeptide linkers may play a critical role in intermodular communication to facilitate the transfer of the growing intermediates between the interacting NRPS and/or PKS modules; (5) post-translational modification of the BLM megasynthetase has been accomplished by a single PPTase with broad carrier protein specificity; and (6) BlmIV/BlmIII-templated assembly of the BLM bithiazole moiety requires intriguing protein juxtaposition and modular recognition. These results lay the foundation to investigate the molecular basis for intermodular communication between NRPS and PKS in hybrid peptide-polyketide natural product biosynthesis and set the stage for engineering novel BLM analogues by genetic manipulation of genes governing BLM biosynthesis.

Investigation of the role of the domain linkers in separate site catalysis by Clostridium symbiosum pyruvate phosphate dikinase

Pyruvate phosphate dikinase (PPDK) catalyzes the reversible reaction: ATP + P(i) + pyruvate <--> AMP + PP(i) + PEP using Mg2+ and NH4+ ions as cofactors. The reaction takes place in three steps, each mediated by a carrier histidine residue located on the surface of the central domain of this three-domain enzyme: (1) E-His + ATP <--> E-His-PP.AMP, (2) E-His-PP.AMP + P(i) <--> E-His-P + AMP + PP(i), (3) E-His-P + pyruvate <--> E-His + PEP. The first two partial reactions are catalyzed at an active site located on the N-terminal domain, and the third partial reaction is catalyzed at an active site located on the C-terminal domain. For catalytic turnover, the central domain travels from one terminal domain to the other. The goal of this work is to determine whether the two connecting linkers direct the movement of the central domain between active sites during catalytic turnover. The X-ray crystal structure of the enzyme suggests interaction between the two linkers that may result in their coordinated movement. Mutations were made at the linkers for the purpose of disrupting the linker-linker interaction and, hence, synchronized linker movement. Five linker mutants were analyzed. Two of these contain 4-Ala insertions within the solvated region of the linker, and three have 3-residue deletions in this region. The efficiencies of the mutants for catalysis of the complete reaction as well as the E-His + ATP <--> E-His-PP.AMP partial reaction at the N-terminal domain and the E-His + PEP <--> E-His-P + pyruvate reaction at the C-terminal domain were measured to assess linker function. Three linker mutants are highly active catalysts at both active sites, and the fourth is highly active at one site but not the other. These results are interpreted as evidence against coordinated linker movement, and suggest instead that the linkers move independently as the central domain travels between active sites. It is hypothesized that while the linkers play a passive role in central domain-terminal domain docking, their structural design minimizes the conformational space searched in the diffusion process.

Stable linker peptides for a cellulose-binding domain-lipase fusion protein expressed in Pichia pastoris

Fusion proteins composed of a cellulose-binding domain from Neocallimastix patriciarum cellulase A and Candida antarctica lipase B were constructed using different linker peptides. The aim was to create proteolytically stable linkers that were able to join the functional modules without disrupting their function. Six fusion variants containing linkers of 4-44 residues were expressed in Pichia pastoris and analysed. Three variants were found to be stable throughout 7-day cultivations. The cellulose-binding capacities of fusion proteins containing short linkers were slightly lower compared with those containing long linkers. The lipase-specific activities of all variants, in solution or immobilized on to cellulose, were equal to that of the wild-type lipase.

Selective protein-protein interactions direct channeling of intermediates between polyketide synthase modules

Polyketide synthases (PKSs) have represented fertile targets for rational manipulation via protein engineering ever since their modular architecture was first recognized. However, the mechanistic principles by which biosynthetic intermediates are sequentially channeled between modules remain poorly understood. Here we demonstrate the importance of complementarity in a remarkably simple, repetitive structural motif within these megasynthases that has been implicated to affect intermodular chain transfer [Gokhale, R. S., et al. (1999) Science 284, 482]. The C- and N-terminal ends of adjacent PKS polypeptides are capped by short peptides of 20-40 residues. Mismatched sequences abolish intermodular chain transfer without affecting the activity of individual modules, whereas matched sequences can facilitate the channeling of intermediates between ordinarily nonconsecutive modules. Thus, in addition to substrate-PKS interactions and domain-domain interactions, these short interpolypeptide sequences represent a third determinant of selective chain transfer that must be taken into consideration in the protein engineering of PKSs. Preliminary biophysical studies on synthetic peptide mimics of these linkers suggest that they may adopt coiled-coil conformations.

Hybrid peptide-polyketide natural products: biosynthesis and prospects toward engineering novel molecules

The structural and catalytic similarities between modular nonribosomal peptide synthetase (NRPS) and polyketide synthase (PKS) inspired us to search for hybrid NRPS-PKS systems. By examining the biochemical and genetic data known to date for the biosynthesis of hybrid peptide-polyketide natural products, we show (1) that the same catalytic sites are conserved between the hybrid NRPS-PKS and normal NRPS or PKS systems, although the ketoacyl synthase domain in NRPS/PKS hybrids is unique, and (2) that specific interpolypeptide linkers exist at both the C- and N-termini of the NRPS and PKS proteins, which presumably play a critical role in facilitating the transfer of the growing peptide or polyketide intermediate between NRPS and PKS modules in hybrid NRPS-PKS systems. These findings provide new insights for intermodular communications in hybrid NRPS-PKS systems and should now be taken into consideration in engineering hybrid peptide-polyketide biosynthetic pathways for making novel "unnatural" natural products.

Cloning, nucleotide sequence, and heterologous expression of the biosynthetic gene cluster for R1128, a non-steroidal estrogen receptor antagonist. Insights into an unusual priming mechanism

R1128 substances are anthraquinone natural products that were previously reported as non-steroidal estrogen receptor antagonists with in vitro and in vivo potency approaching that of tamoxifen. From a biosynthetic viewpoint, these polyketides possess structurally interesting features such as an unusual primer unit that are absent in the well studied anthracyclic and tetracyclic natural products. The entire R1128 gene cluster was cloned and expressed in Streptomyces lividans, a genetically well developed heterologous host. In addition to R1128C, a novel optically active natural product, designated HU235, was isolated. Nucleotide sequence analysis of the biosynthetic gene cluster revealed genes encoding two ketosynthases, a chain length factor, an acyl transferase, three acetyl-CoA carboxylase subunits, two cyclases, two oxygenases, an amidase, and remarkably, two acyl carrier proteins. Feeding studies indicate that the unusual 4-methylvaleryl side chain of R1128C is derived from valine. Together with the absence of a dedicated ketoreductase, dehydratase, or enoylreductase within the R1128 gene cluster, this suggests a functional link between fatty acid biosynthesis and R1128 biosynthesis in the engineered host. Specifically, we propose that the R1128 synthase recruits four subunits from the endogenous fatty acid synthase during the biosynthesis of this family of pharmacologically significant natural products.

Dipeptide formation on engineered hybrid peptide synthetases

BACKGROUND

Nonribosomal peptide synthetases (NRPSs) are modular 'megaenzymes' that catalyze the assembly of a large number of bioactive peptides using the multiple carrier thiotemplate mechanism. The modules comprise specific domains that act as distinct units to catalyze specific reactions associated with substrate activation, modification and condensation. Such an arrangement of biosynthetic templates has evoked interest in engineering novel NRPSs.

RESULTS

We describe the design and construction of a set of dimodular hybrid NRPSs. By introducing domain fusions between adenylation and thiolation (PCP) domains we designed synthetic templates for dipeptide formation. The predicted dipeptides, as defined by the specificity and arrangement of the adenylation domains of the constructed templates, were synthesized in vitro. The effect of the intramolecular fusion was investigated by determining kinetic parameters for substrate adenylation and thiolation. The rate of dipeptide formation on the artificial NRPSs is similar to that of natural templates.

CONCLUSIONS

Several new aspects concerning the tolerance of NRPSs to domain swaps can be deduced. By choosing the fusion site in the border region of adenylation and PCP domains we showed that the PCP domain exhibits no general substrate selectivity. There was no suggestion that selectivity of the condensation reaction was biased towards the donor amino acid, whereas at the acceptor position there was a size-determined selection. In addition, we demonstrated that a native elongation module can be converted to an initiation module for peptide-bond formation. These results represent the first example of rational de novo synthesis of small peptides on engineered NRPSs.