Exploring the chemical diversity of environmental actinobacteria using genome mining and synthetic biology

Abstract

Natural products derived from biosynthetic gene clusters (BGCs) in microorganisms are vital for medicine, agriculture, and industry. However, rediscovering known compounds and the resource-intensive nature of drug development delays the discovery of novel compounds. This study aimed to address these challenges by integrating genome mining and synthetic biology to explore the chemical diversity of actinobacteria from the oligotrophic environment of Cuatro Ciénegas, Mexico. In the first chapter of this work, we evaluate the genomic and biosynthetic potential of the environmental strains Lentzea sp. CC55 and Actinokineospora sp. PR83 isolated from Cuatro Ciénegas environment. Comparative genomics revealed open pan-genomes comprising 568 and 965 unique genes, respectively. BGC similarity networks identified unique clusters, including terpenes, RiPPs, NRPS, and polyketides. Both strains demonstrated antimicrobial and cytotoxic activities. Lentzea sp. CC55 culture showed cytotoxic activity only in liquid cultures while Actinokineospora sp. PR83 presented activity against B. subtilis for solid media. The Biosynthetic Novelty Index (BiNI) confirmed these strains as high-priority candidates for novel natural product discovery. In the second chapter, we applied the BiNI index in our collection of strains from Cuatro Ciénegas to select the candidates with the highest novelty. From the selected candidates we standardized cloning strategies for BGC expression in heterologous hosts. Using Golden Gate assembly and CRISPR/Cas9-based DNA assembly, genes were domesticated, assembled into transcriptional units, and transferred to Streptomyces albidoflavus UO-FLAV-004. Challenges with toxic fragments and incomplete assemblies were resolved using optimized host strains, resulting in the successful conjugation of core modules and the initiation of heterologous expression. In the third chapter, we explore different tools of metabolomics for the detection of metabolites produced by engineered strains. For NRPS the prediction indicates a theoretical mass: 495 Da. The RiPP corresponds to a lanthipeptide type I with a theoretical mass of 2318.9 Da. Molecular network analysis and dereplication only identified compounds produced intrinsically encoded by the host. This research presents a genomic-guide pipeline for exploring the biosynthetic potential of Cuatro Ciénegas actinobacteria, providing insights into the BGC prioritization and application of synthetic biology techniques to improve the intricate pipelines for new chemical diversity in Natural Products research.