The project NAfrAM started with preliminary data of the four prioritized samples from four different Antarctic ecosystems (lake, stream, wetland, pond). The two samples were vertically stratified microbial mats and two samples were non-structured microbial mats. The preliminary data consists of 16S amplicon sequencing and shotgun metagenomic sequencing data prepared by Illumina short-read sequencing technology. During the first three months of the project, I extracted high molecular weight DNA from these samples and the samples were also sequenced by the Oxford Nanopore Technologies method to produce long-read sequencing data. Such a complex ultra-deep sequencing dataset was used to produce high-quality metagenomic assemblies where the long-read data were assembled first and then corrected by short-read Illumina data.
The resulting metagenomic assemblies were used for comprehensive bioinformatic analysis. The bioinformatics analysis consists of taxonomic classification of metagenomic contigs, BGCs detection, BGC taxonomic classification, BGCs clustering to gene cluster families (GCFs) to allow more accurate evaluation of biosynthetic potential and BGCs diversity, binning of metagenomic contigs into metagenome-assembled genomes (MAGs), linking the BGCs with MAGs, phylogenetic and taxonomic analysis of MAGs, analysis of the environmental distribution of BGCs, semi-manual identification of precursor peptides for RiPPs BGCs, prioritization of nearly fifty BGCs possibly producing antimicrobial compound. Further, the datasets of more than 6,000 BGCs detected across 4 Antarctic samples, more than 1,000 high-quality (HQ) and medium-quality (MQ) MAGs, and nearly 370 precursor peptides of diverse RiPPs pathways belong to the main achievements of the bioinformatic part. Such datasets constitute invaluable data for further data mining studies that can help in the investigation of Antarctic biodiversity and understanding of microbial evolution in Antarctica. Importantly, these data can directly serve the biosynthesis or chemical synthesis of specialized metabolites with pharmaceutical or biotechnological potential and help to understand their biological functions.
The bioinformatic part informed the experimental part which builds on prioritized BGCs possibly producing antimicrobial compounds. In this direction, ten novel BGCs from Antarctic microbes were further selected for the experimental part, namely 1x thiopeptide, 1x hybrid lanthipeptide-NRPS-PKS-Terpene, 1x hybrid NRPS-PKS, 3x lasso peptide, 3x lanthipeptide, and 1x hybrid homoserine lactone-lasso peptide. All BGCs were successfully PCR amplified from Antarctic microbial communities and assembled into expression vectors. Considering the project's timescale, only one lasso peptide and three lanthipeptide BGCs, containing unique features in precursor peptides (polyarginine regions) were selected for detailed experimental investigations. In the first part, the lasso peptide BGCs were further optimized to increase the chance for successful heterologous expression on the level of promoter, RBS and codon-optimization for Escherichia coli. After sequencing verification of expression plasmids with inserted BGCs, lasso peptide and lanthipeptide BGC were heterologously expressed in Escherichia coli BL21(DE3). The BGCs were expressed under several experimental setups (temperature 20°Cfor 72 hours/37°C for 24 hours, lasso peptide in M9 media, lanthipeptide in LB media. each in triplicate). However, neither the lasso peptide nor the lanthipeptide final product was detected using LC-ESI-MS. Thus, the next step was heterologous expression and purification of every single protein involved in biosynthesis. These experiments revealed soluble expression for all proteins from lanthipeptide BGCs; however, it revealed insoluble expression of peptidase from lasso peptide BGCs responsible for cleavage of the leader peptide from the core peptide in precursor peptide, a step necessary for cyclization resulting in a final matured product. Therefore, these experiments directed the effort to the lanthipeptide BGCs. All single proteins were purified and used for in vitro reaction. Although the final product of the in vitro reactions was not detected by LC-ESI-MS, the final product was detected by MALDI-TOF MS. It also suggested that original experiments with complete BGCs also resulted in the production of the final product, but the native product is hard to detect using LC-ESI-MS.
