Mosaic Ends Tagmentation (METa) assembly for extremely efficient construction of functional metagenomic librarie

Abstract

Functional metagenomic libraries, physical bacterial libraries which allow the high-throughput capture and expression of microbiome genes, have been instrumental in the sequence-naive and cultivation-independent discovery of novel genes from microbial communities. Preparation of these libraries is limited by their high DNA input requirement and their low cloning efficiency. Here, we describe a new method, METa assembly, for extremely efficient functional metagenomic library preparation. We apply tagmentation to metagenomic DNA from soil and gut microbiomes to prepare DNA inserts for high-throughput cloning into functional metagenomic libraries. The presence of mosaic end sequences in the resulting DNA fragments synergizes with homology-based assembly cloning to result in a 300-fold increase in library size compared to traditional blunt cloning based protocols. Compared to published libraries prepared by state-of-the-art protocols we show that METa assembly is on average 23- to 270-fold more efficient and can be effectively used to prepare gigabase-sized libraries with as little as 200 ng of input DNA. We demonstrate the utility of METa assembly to capture novel genes based on their function by discovering novel aminoglycoside (26% amino acid identity) and colistin (36% amino acid identity) resistance genes in soil and goose gut microbiomes. METa assembly provides a streamlined, flexible, and efficient method for preparing functional metagenomic libraries, enabling new avenues of genetic and biochemical research into low biomass or scarce microbiomes. IMPORTANCEMedically and industrially important genes can be recovered from microbial communities by high-throughput sequencing but are limited to previously sequenced genes and their relatives. Cloning a metagenome en masse into an expression host to produce a functional metagenomic library is a sequence-naive and cultivation-independent method to discover novel genes. This directly connects genes to functions, but the process of preparing these libraries is DNA greedy and inefficient. Here we describe a library preparation method that is an order of magnitude more efficient and less DNA greedy. This method is consistently efficient across libraries prepared from cultures, a soil microbiome, and from a goose fecal microbiome and allowed us to discover novel antibiotic resistance genes. This new library preparation method will potentially allow for the functional metagenomic exploration of microbiomes that were previously off limits due to their rarity or low microbial biomass, such biomedical swabs or exotic samples.