Long-read based de novo assembly of low-complexity metagenome samples results in finished genomes and reveals insights into strain diversity and an active phage system

Abstract

Background Complete and contiguous genome assemblies greatly improve the quality of subsequent systems-wide functional profiling studies and the ability to gain novel biological insights. While a de novo genome assembly of an isolated bacterial strain is in most cases straightforward, more informative data about co-existing bacteria as well as synergistic and antagonistic effects can be obtained from a direct analysis of microbial communities. However, the complexity of metagenomic samples represents a major challenge. While third generation sequencing technologies have been suggested to enable finished metagenome-assembled genomes, to our knowledge, the complete genome assembly of all dominant strains in a microbiome sample has not been demonstrated. Natural whey starter cultures (NWCs) are used in cheese production and represent low-complexity microbiomes. Previous studies of Swiss Gruyère and selected Italian hard cheeses, mostly based on amplicon metagenomics, concurred that three species generally pre-dominate: Streptococcus thermophilus , Lactobacillus helveticus and Lactobacillus delbrueckii . Results Two NWCs from Swiss Gruyère producers were subjected to whole metagenome shotgun sequencing using the Pacific Biosciences Sequel and Illumina MiSeq platforms. In addition, longer Oxford Nanopore Technologies MinION reads had to be generated for one to resolve repeat regions. Thereby, we achieved the complete assembly of all dominant bacterial genomes from these low-complexity NWCs, which was corroborated by a 16S rRNA amplicon survey. Moreover, two distinct L. helveticus strains were successfully co-assembled from the same sample. Besides bacterial chromosomes, we could also assemble several bacterial plasmids and phages and a corresponding prophage. Biologically relevant insights were uncovered by linking the plasmids and phages to their respective host genomes using DNA methylation motifs on the plasmids and by matching prokaryotic CRISPR spacers with the corresponding protospacers on the phages. These results could only be achieved by employing long-read sequencing data able to span intragenomic as well as intergenomic repeats. Conclusions Here, we demonstrate the feasibility of complete de novo genome assembly of all dominant strains from low-complexity NWCs based on whole metagenomics shotgun sequencing data. This allowed to gain novel biological insights and is a fundamental basis for subsequent systems-wide omics analyses, functional profiling and phenotype to genotype analysis of specific microbial communities. Electronic supplementary material The online version of this article (10.1186/s12866-019-1500-0) contains supplementary material, which is available to authorized users.

An integrated model system to gain mechanistic insights into biofilm formation and antimicrobial resistance development inPseudomonas aeruginosaMPAO1

“…In addition, a 2 x 300 bp paired end library (Illumina Nextera XT DNA kit) was sequenced on a MiSeq. Polishing of the assembly with Illumina reads, circularization, start alignment using dnaA and final verification of assembly completeness were performed as described previously [64]. The quality of the aligned reads and the final chromosome was assessed using Qualimap [65].…”

metaVaR: introducing metavariant species models for reference-free metagenomic-based population genomics

“…These approaches were successfully applied on prokaryotes (Sedlar et al, 2016;Parks et al, 2017;Stewart et al, 2019) but the main limit of the alignment-based approach still relies on the availability, completeness and quality of reference genomes that can be constructed from metagenomic data (Parks et al, 2017). The approach took recently advantages of long-read sequencing (Huson et al, 2018;Somerville et al, 2019). However, due to the large genome size of eukaryotes and the difficulties to obtain high molecular weight DNA, yet, such approaches did not produce results on eukaryotes.…”

Near-complete Lokiarchaeota genomes from complex environmental samples using long and short read metagenomic analyses

“…Although several long-read assembly methods have been successfully used for single-genome de novo assembly Koren et al, 2017) and new promising tools are being developed for metagenomic assembly (Kolmogorov et al, 2019a), how well these methods perform with real metagenomic data still remains to be properly evaluated. Yet, an increasing number of long read-based metagenomics studies demonstrate that it is possible to reconstruct complete or near-complete genomes from mock communities and natural microbial communities of varying complexity (Bertrand et al, 2019;Hao et al, 2018;Moss and Bhatt, 2018;Nicholls et al, 2019;Somerville et al, 2019).…”

“…In metagenomes, repetitive sequences can be present within a single genome, as well as within several genomes that share similar regions (e.g., closely related strains, horizontally transferred genes). Therefore, if reads are long enough to span across repetitive regions, and also cover regions that are unique to a particular genome, their inclusion in metagenome assemblies can help to separate the genomic sequences of different organisms, reducing the number of chimeric contigs (Somerville et al, 2019). Furthermore, provided enough sequencing depth, metagenomic assemblies of long-read sequence datasets are expected to generate longer contigs than short-read-based assemblies, reducing potential risks of misclassifying contigs in metagenome binning efforts.…”