Deciphering and engineering of the molecular diversity of NRPS-derived azacyclic alkaloids in bacteria (NRPSBacAza) – NRPSBacAza

We employ a multidisciplinary approach, including microbiology, chemoinformatics, untargeted metabolomics, biochemistry, synthetic biology and analytical chemistry, to achieve our objectives.

This study firmly elucidates the biosynthesis of azetidomonamide A, a cyclocarbamate produced by P. aeruginosa. The post-NRPS/BVMO steps include ketoreduction, dehydration and stereospecific hydroxylation. A one-pot reaction with all biosynthetic enzymes enabled the production of azetidomonamide A in vitro. Bioinformatics, structural and mechanistic analyses of the condensation domain-like dehydratase was performed, providing valuable insights into the non-canonical functions performed by NRPS domains.

Secondly, the biosynthetic pathway of pyrrolizwillines was identified in X. hominickii. A key hydrolase at the branched point that converts the common pathway intermediate pyrrolizixenamide to pyrrolizwilline was biochemically and structurally characterized. Remarkably, the formation of pyrrolizwillines requires a step of non-enzymatic condensation with a primary metabolite.

Previous studies have shown that bacterial PA pathways in general display a large metabolic plasticity, owing to the substrate promiscuity of the NRPS adenylation domains. Thus, a coexpression strategy was applied for product engineering in E. coli. The NRPS/BVMO pair involved in the biosynthesis of the PA pyrrolizixenamide was coexpressed with AzeJ, an enzyme from the azetidomonamide pathway that produces the four-membered NRPS precursor (i. e. azetidine 2-carboxylic acid (AZC)). This allowed the production of new-to-nature AZC-incorporated pyrrolixenamides.

This study deepens current understanding of the metabolic plasticity and biosynthesis of bacterial PA and related compounds, and highlight the importance of non-enzymatic reactions in chemical diversification in these pathways. Moreover, we reveal valuable structural and mechanistic insights into the non-canonical function of NRPS condensation domains, which significantly advances our understanding of NRPS enzymology.

There are many perspectives. In a short term, azetidomonamide biosynthesis represents an excellent model to study other non-canonical activities within NRPS domains. Moreover, leveraging the obtained biosynthetic knowledges together with genome mining will allow to access new bacterial PA and related compounds, in particular from pathogens and those of ecological importance. In a long term, we will tackle the question of the biological function of these compounds in bacteria by an integrative approach including methods from chemical biology, -omics, microbiology and biochemistry.

While many biosynthetic pathways in bacteria lead to one or a limited number of products, certain systems produce a diversity of metabolites. The pathways to the bacterial pyrrolizidine alkaloids (PAs) represent striking examples of this phenomenon. A common biosynthetic scheme involving a bimodular nonribosomal peptide synthetase (NRPS) and a monooxygenase, in combination with accessory enzymes and/or spontaneous chemical reactions, leads to diverse azacycles such as pyrrolizidines and cyclocarbamates. In recent work, several of these compounds have been found to attenuate virulence of the producing bacteria during interaction with its host. The molecular mechanisms governing such diversification and underlying their biological activity remain unknown. Given the high interest and broad pharmaceutical potential of azacycle-containing molecules, this project proposes to decipher the molecular basis for the chemical diversity of pyrrolizidine and related alkaloids in bacteria, to elucidate structure-function relationships for key enzymes in the pathways, and to apply the obtained knowledge to generate new-to-nature derivatives with interesting bioactivity, notably as anti-virulence compounds of a major human pathogen, Pseudomonas aeruginosa.