Natural products serve as essential scaffolds for pharmaceuticals, agrochemicals, food additives, and cosmetics. The practical application of natural products, however, often requires precise structural modifications of the compounds. Synthetic biology offers a sustainable route to engineer these complex molecules while optimizing them for desired applications. Thiotemplated enzymes, like polyketide synthases or on-ribosomal peptide synthetases, are particularly promising for the synthetic biology approach, as they function in a modular, factory-like manner. The natural products formed are used in numerous pharmaceutical and other applications, enabled by their high structural diversity. To achieve such structural variations, several strategies are employed by the megasynthases, for example, the usage of dedicated building blocks with defined reactivities. In addition, the condensation of independent polyketide chains serves as an effective mechanism to enhance molecular complexity. To exploit the potential of these enzymatic strategies, a comprehensive understanding of the underlying biocatalysts is necessary. Until recently, enzymes capable of mediating polyketide chain fusion had been characterized only among a few free-standing enzymes. Two individual modular biocatalysts predicted to catalyze such fusions were identified in the biosynthetic pathways of ikarugamycin and malleicyprol. In ikarugamycin biosynthesis, an amino acid serves as a linker between polyketide building blocks, whereas the malleicyprol pathway suggested a direct fusion of two polyketide chains. Therefore, studying the malleicyprol system provided an opportunity to elucidate a previously unknown mode of biocatalysis that could be harnessed to design custom natural products. Further, it is important to note that malleicyprol itself is a virulence factor produced by pathogenic bacteria, such as those responsible for causing the tropical disease melioidosis. Studying the enzymes involved in its biosynthesis not only offered insights into novel catalytic strategies but could also identify targets for specific inhibition that would be finally used as therapeutics. Based on the results of this dissertation, a model for the complete biosynthetic pathway to malleicyprol and its inactivation is described. The final steps of the biosynthesis are catalyzed by the malleicyprol synthase BurF, a trans-AT PKS-NRPS hybrid modular enzyme consisting of two polyketide modules at the N-terminus, and an unusual C-terminal NRPS-like module. The N-terminal ACP1 is loaded with an octanoyl adenylate intermediate, which the FAAL BurM generates by condensing octanoic acid with ATP (Manuscript A). Additionally, the homologous fatty acid decanoic acid is activated to some extent by BurM and utilized from the Summary bur assembly line, resulting in various malleicyprol congeners with altered side chains (Manuscript A). After the fatty acids are loaded onto the ACP1, they are elongated twice with malonyl-CoA by the polyketide modules to produce a bound fatty acid-polyketide intermediate bound on ACP3 (Manuscript A). Based on the structure of malleicyprol,103 the previous discovery of trigonic acid,106 and the anticipated domain architecture of BurF, it was assumed that the fatty acid-polyketide intermediate is then merged with T domain-bound trigonic acid. This building block contains the cyclopropanol motif and is an α-hydroxy acid, indicating that it must be activated to be loaded onto the assembly line. However, the C-terminal NRPS-like module of BurF lacks any of the previously known domains for α-hydroxy acid activation. Therefore, it has remained elusive how trigonic acid should be introduced into the bur assembly line. By genetic manipulations and in vitro characterization of the free-standing adenylation-thiolation didomain protein BurJ, the entry point of trigonic acid loading was elucidated in Manuscript B. Discovering this enzyme represents the first identification of a bacterial adenylation domain known to directly activate an α-hydroxy acid, which had only been known from few fungal adenylation domains.40 From the BurJ T domain, bound trigonic acid is transferred to the BurJ partner protein BurH. This protein belongs to the well-characterized family of FkbH-like proteins that usually shuttle the primary metabolite 1,3-bisphosphoglycerate into natural product assembly lines.79 BurH differs in its activity from all other members of the family by processing the building block trigonic acid. This change in activity is also reflected in the BurH primary sequence: although the protein contains a conserved, catalytic cysteine residue for trigonic acid loading in its acyltransferase domain, the second phosphatase domain is altered compared to other members of the FkbH-like protein family. From the BurH cysteine residue, the trigonic acid building block is passed onto the BurF T domain which was demonstrated in vitro, by genetic inactivation of the burH gene, and by genetic complementation.
Jonas Fiedler (Sat,) studied this question.