Bioinformatics applied to natural products discovery processes: systematization, biosynthetic evidences, and isolation of promising species

Abstract

Genome-guided strategies were applied to examine Betaproteobacteria species potential for the biosynthesis of nonribosomal peptides. A generalizable strategy was created to track similarities in enzymatic reactions of nonribosomal peptides synthetases in order to organize their capability of assembling monomers building the peptides backbones. Databases and user-friendly software were adopted making this strategy a comprehensive one. Databases and software adopted, as well as, NCBI, KEGG, NORINE, antiSMASH, Cystoscape, Gitools, MEGA e Clustal were used for this purpose. Betaproteobacteria species showed to possess biosynthetic similarities in assembling monomers for the peptide backbone of a nonribosomal peptide. These evidences were correlated giving similarities indexes between species and their distribution between similar genomes. Predictions were fragmented in several ways, for example, monomers, pairs and triads. Correlation analyses displayed that pairs it is the best way of tracking similarities. This result turned possible to create a strategy, named XPAIRT (eXPAndable Identification of amino acids in nonRibosomal peptides Tendencies) correlating pairs of peptides and their similar genomes via Jaccard Index and phylogeny. Thought these investigations it was noticed that Betaproteobacteria species generally assemble asp.orn and orn.ser, mainly Burkholderia species, among other pairs of peptides. Further analysis showed that species from the genera Burkholderia are the most promising ones due to their Biosynthetic Gene Cluster counting for all available Betaproteobacteria genomes. These species were further analyzed and a standard strain, Burkholderia thailandensis, was used to the identification of intraspecific variation for their biosynthetic potential. A specific study on Biosynthetic Gene Cluster variation was proceeded for discovering disparities between chromosomes 1 and 2, and a standard antibiotic producer strain, S. coelicolor. Results showed that B. thailandensis have different possibilities for biosynthesizing natural products. Even thought, common classes of compounds such as, Terpenes, Bacteriocins, T1PKS and Nonribosomal Peptides were identified for all strains. As Burkholderia species were the main target in this work, a genome-guided method was developed for isolating as much strains as possible from environmental samples. This very method took into account the basic needs for a microorganism to survive: a) the type of microbiome that microorganisms of interest coexist, analyzed through metagenomics, b) resistance to antibiotics and metals, c) ability to metabolize compounds with biological role, d) cell growth related to different nutrients, and e) cell growth under pH variations. The strategy was successful for diverse types of samples. These exceptional experiments are part of a novel way of working with Natural Products, using genomic, bioinformatics and visual statistical analysis in order to access common characteristics and uniqueness of species guiding the search of medically relevant natural products.