Thesis advisor: Steven D. Bruner / Thesis advisor: Larry W. McLaughlin / Nonribosomal peptides are therapeutically important natural products produced through pathways that utilize large multimodular enzymes, termed nonribosomal peptide synthetases (NRPSs). Central to the assembly line methodology, the monomer building blocks and the growing polymer chain are covalently linked to dedicated peptidyl carrier protein domains as phosphopantetheinyl thioesters. Although structures of multidomain NRPS fragments have been solved recently, the active conformation of the carrier domains with their attached phosphopantetheinyl arms has not been determined. Significant conformational changes in carrier domains are likely to occur as the domains shuttle peptidyl phosphopantetheinyl thioesters between the active sites of the partner domains. This thesis focuses on the application of the synthetic isosteric non-hydrolyzable CoA analogs to manipulate carrier domain geometry of NRPS assemblies through. The synthetic conjugates are designed to deliver an inhibitor moiety to a domain of interest. Using this strategy, various complexes have been designed to direct the phosphopantetheinyl arm to active sites of adenylation domains and thioesterase domains in catalytically relevant conformations. The structurally restrained multidomain NRPS assemblies are useful for elucidating the complex structure and mechanism of NRPSs. An X-ray crystal structure of a peptidyl carrier-thioesterase NRPS didomain fragment from enterobactin synthetase has been solved with a phosphopantetheinyl analog which forms a cross-link between the two domains. This structure provides, for the first time, detailed insights into the phosphopantetheinyl arm interaction with an NRPS partner domain, as well as an active confirmation of a mutidomain NRPS in the holo-form. In addition, the hydrolytically stable CoA analogs have been successfully used as probes in the structural and mechanistic study of a CoA-utilizing enzyme DpgC, a unique cofactor-independent dioxygenase involved in vancomycin biosynthesis. / Thesis (PhD) — Boston College, 2009. / Submitted to: Boston College. Graduate School of Arts and Sciences. / Discipline: Chemistry.

Alternaria brassicicola is a necrotrophic pathogen that causes black spot disease on virtually all cultivated Brassicas, A. brassicicola is renowned for its ability to prodigiously produce secondary metabolites. To test the hypothesis that secondary metabolites produced by A. brassicicola contribute to pathogenicity, we identified seven nonribosomal peptide synthetases (NPSs) and 10 polyketide synthases (PKSs) in the A. brassicicola genome. The phenotype resulting from knockout mutations of each PKS and NPS gene was investigated with an emphasis on discovery of fungal virulence factors. A highly efficient gene disruption method using a short linear double stranded DNA construct with minimal elements was developed, optimized, and used to functionally disrupt all NPS and PKS genes in A. brassicicola. Three NPS and two PKS genes, and one NPS-like gene appeared to be virulence factors based upon reduced lesion development of each mutant on inoculated green cabbage and Arabidopsis compared with the wild-type strain. Furthermore some of the KO mutants exhibited developmental phenotypic changes in pigmentation and conidiogenesis. To further characterize the roles of several genes of interest in A. brassicicola development and pathogenesis, the genes AbNPS2, AbPKS9, and NPS-like tmpL were selected for in-depth functional analysis. We provide substantial evidence that the AbNPS2-associated metabolite is involved in conidial cell wall construction, possibly as an anchor connecting two cell wall layers. We also characterized a biosynthetic gene cluster harboring the AbPKS9 gene and demonstrated that this cluster is responsible for the biosynthesis of depudecin, an inhibitor of histone deacetylases and a minor virulence factor. Finally, we demonstrated that a NPS-like protein named TmpL is involved in a filamentous fungi-specific mechanism for regulating levels of intracellular reactive oxygen species during conidiation and pathogenesis in both plant and animal pathogenic fungi. Collectively our results indicate that small molecule nonribosomal peptides and polyketides in A. brassicicola play diverse, but also fundamental, roles in fungal development and pathogenesis. / Ph. D.

La colistine, antibiotique appartenant à la famille des polymyxines, est un polypeptide cyclique, cationique, ciblant les membranes bactériennes. Elle est produite par Paenibacillus polymyxa via des complexes multi-enzymatiques appelés Non-Ribosomal Peptides Synthétases (NRPS). Dans le cas de la mucoviscidose, et malgré des effets secondaires importants, la colistine est utilisée comme ultime recours pour lutter contre les bactéries Gram-négatives multirésistantes responsables d’infections pulmonaires dont Pseudomonas aeruginosa. Jusqu’ici les systèmes génétiques à l’origine de la production de la colistine étaient peu connus. Au cours de cette étude, nous avons caractérisé par LC-MS haute résolution des molécules antimicrobiennes, dont des colistines, produites par un nouveau Paenibacillus. Afin d’identifier et de cloner le cluster de gène responsable de la production de ces antibiotiques, une banque d’ADN génomique a été construite et criblée par homologie de séquence avec des systèmes de production déjà connus. Ce criblage a permis de sélectionner quatre clones d’intérêt. L’étude in silico de leurs séquences a permis d’identifier les différents modules d’un nouveau cluster NRPS qui serait à l’origine de la synthèse de variants de la colistine. À terme, cette découverte pourrait permettre de mieux contrôler la production de la colistine et d’obtenir des composés plus actifs et/ou présentant des effets secondaires amoindris. En parallèle à ce premier travail, nous avons également recherché la présence de nouvelles NRPS chez une centaine de micro-organismes issus d’une station d’étude environnementale du laboratoire (vasière intertidale). Ce travail a permis de découvrir des nouvelles séquences et d’isoler un nouveau micro-organisme producteur d’antibiotique(s). / Colistin is a cationic cyclic polypeptide antibiotic belonging to the polymyxin family and targeting bacterial membranes. It is produced by Paenibacillus polymyxa through a Nonribosomal Peptide Synthetase (NRPS) mechanism. In the context of cystic fibrosis (CF), colistin is used for the treatment of lung infections caused by multiresistant Gram-negative bacteria including Pseudomonas aeruginosa. Unfortunately, this molecule is also known for its strong side effects. So far, genetic systems controlling the production of polymyxins were little known. In this study we characterized by High-resolution LC-MS the antimicrobial molecules, including colistins, of a new Paenibacillus. A genomic library of this strain was constructed and screened to identify genes involved in the production of these antibiotics. A degenerated PCR screening was performed and allowed to select four clones in the genomic library. In silico study allowed to identify a new NRPS gene cluster responsible for the biosynthesis of colistin variants. In the future, this work might allow the harnessing of the production of colistin derived structures, more active and/or showing fewer side effects. In parallel, a second investigation was performed in order to find new NRPS genes in a collection of one hundred intertidal mudflat bacterial isolates. This work has allowed the identification of new sequences and the characterization of a new antimicrobial producing strain.

Modulare Polyketidsynthasen (PKS) sind Multienzymsysteme des bakteriellen Sekundärstoffwechsels. An ihnen läuft eine schrittweise Biosynthese vielfältiger Kohlenstoff-Gerüste ab, die von einfachen Carbonsäure-Einheiten ausgeht. Polyketid-Verbindungen zeigen eine große Bandbreite pharmazeutisch interessanter Aktivitäten. In dieser Arbeit wurde eine Reihe von Evolutionsstudien durchgeführt. Zunächst wurden die phylogenetischen Beziehungen zwischen modularen PKS und anderen PKS-Systemen sowie Fettsäuresynthasen untersucht, wodurch ihre zentrale Stellung innerhalb eines langen Evolutionsprozesses gezeigt werden konnte. Eine detaillierte Analyse der Phylogenien von Domänen bakterieller modularer PKS ergab, dass das Ausmaß an Genduplikationen, Genverlusten und Ereignissen horizontalen Gentransfers zwischen den verschiedenen Bakteriengruppen beträchtlich variiert. Aus der Genomsequenz des Actinobakteriums Streptomyces avermitilis wurden die Phylogenien aller Domänentypen rekonstruiert. Der Vergleich dieser Einzelphylogenien ermöglichte es, eine Reihe von homologen Rekombinationsereignissen aufzufinden. Homologe Rekombination scheint der Hauptmechanismus zu sein, auf dem die Strukturvielfalt der Polyketide in Bakterien beruht. Mit Hilfe eines „genome mining“-Ansatzes konnte im Genom des Cyanobakteriums Nostoc punctiforme eine Reihe von Biosynthese-Clustern, die zu den PKS und nichtribosomalen Peptidsynthetasen gehören, identifiziert werden. Durch chromatographische und massenspektrometrische Analysen von Zellextrakten und Kulturüberständen konnten einige der Biosynthese-Cluster bestimmten Metaboliten zugeordnet werden. Eines der Cluster wurde hinsichtlich des produzierten Metaboliten und der Regulationsstruktur eingehender charakterisiert. Die Folgerungen aus den gewonnen Ergebnissen werden im allgemeinen Zusammenhang der Evolution metabolischer Diversität ausführlich diskutiert. / Modular polyketide synthases (PKS) are multienzym systems of bacterial secondary metabolism. They perform a stepwise biosynthesis of diverse carbon skeletons from simple carboxylic acid units. Polyketide compounds possess a wide range of pharmaceutically interesting activities. In this study, a series of evolutionary analyses was performed. Initially, the phylogenetic relationships between modular PKS and other PKS systems as well as fatty acid synthases were investigated revealing their central position within a long evolutionary process. In detail reconstruction of the phylogenies of bacterial modular PKS domains demonstrated that the extent of gene duplications, gene losses and horizontal gene transfer events varies considerably between different bacterial groups. Using the genome sequence of the actinobacterium Streptomyces avermitilis the phylogenies of all domain types were reconstructed. Comparison of these phylogenies allowed for detecting numerous events of homologous recombination, which appears to be the main mechanism underlying polyketide structural diversity in bacteria. A genome mining approach revealed a number of biosynthesis clusters of the PKS and nonribosomal peptide synthetase type in the genome of the cyanobacterium Nostoc punctiforme. Cell extracts and culture supernatants were analysed by means of liquid chromatography and mass spectrometry and some of the biosynthesis clusters could be assigned to specific metabolites. One of the clusters was characterised in greater detail regarding the produced metabolite and the cluster’s regulatory structure. The implications of the results are extensively discussed within the general context of the evolution of metabolic diversity.