Development of methods for the culture-independent discovery of natural products from soil metagenomes

Abstract

The diverse bioactivities of microbial secondary metabolite natural products enabled their use as medical drugs for decades. Especially antibiotics produced by soil bacteria proved to be indispensable for the successful treatment of bacterial infectious diseases. However, the rise of multi-drug resistant bacteria resulting from the ongoing misuse and overuse of antibiotics is currently causing an increasing number of deadly infections, yet the clinical approval rate of new antimicrobial drugs has decreased in the recent past, which is why there is a high demand for new antibiotics. While culture-based natural product discovery from predominantly soil bacteria has so far provided the majority of antibiotics, it nowadays struggles with high rediscovery rates of already known compounds. Since this is partly due to the uncultivable nature of the majority of soil bacteria under conventional laboratory conditions, culture- independent metagenomic natural product discovery approaches provide a promising solution to access this previously missed biosynthetic capacity. However, for that purpose, the natural product encoding biosynthetic gene clusters (BGCs) need to be recovered from soil metagenomes to enable subsequent heterologous expression, which often consists of a laborious and time-consuming process. In this thesis, an efficient strategy for the rapid identification and recovery of complete natural product BGCs from soil metagenomes was developed. BGCs were recovered from a soil metagenome of the Schönbuch forest, which enabled subsequent heterologous expression experiments aiming at the production of the encoded molecules. As a first step of the overall process, three different soil types sampled from the Schönbuch forest were investigated for their biosynthetic potential via amplicon and Illumina shotgun sequencing, which revealed a huge biosynthetic capacity of all three soils. High-quality, high molecular weight (HMW) metagenomic DNA was isolated from one of the soil types to construct a metagenomic fosmid library of 83700 clones that was subsequently sequenced using Illumina and Nanopore technologies. Short- and long-reads were used in a hybrid assembly approach to generate contigs that were analyzed to detect complete BGCs. The method single Nanopore read cluster mining (SNRCM) was developed to identify complete BGCs on single fosmids by aligning the detected BGCs directly to Nanopore long-reads. Using SNRCM, two lasso peptide BGCs were recovered and tested in various heterologous expression experiments. Additionally, the corresponding HMW metagenomic DNA was sequenced, which enabled the direct amplification and recovery of a complete lasso peptide BGC via PCR and subsequent cloning into an expression vector, followed by heterologous expression experiments. Furthermore, using both the sequencing data of the metagenomic DNA as well as the corresponding fosmid library enabled the assembly and recovery of a nonribosomal peptide synthetase (NRPS) BGC from three distinct fosmids of the library that was subsequently transferred to various heterologous hosts aiming at the production of the encoded compound. Overall, this thesis contributes to facilitate and accelerate natural product discovery from soil metagenomes, which can potentially lead to the isolation of new medically relevant compounds such as antibiotics in the near future.