Retrieval of a million high-quality, full-length microbial 16S and 18S rRNA gene sequences without primer bias

• Published in: Nature biotechnology Vol. 36; no. 2; pp. 190 - 195
• Main Authors: Karst, Søren M, Dueholm, Morten S, McIlroy, Simon J, Kirkegaard, Rasmus H, Nielsen, Per H, Albertsen, Mads
• Format: Journal Article
• Language: English
• Published: New York Nature Publishing Group US 01.02.2018; Nature Publishing Group
• Subjects: 45/23; 45/77; 45/90; 631/114/1386; 631/326/2565/2134; 631/326/2565/855; 704/158/2452; 704/158/855; Agriculture; Archaea; Archaea - classification; Archaea - genetics; Bacteria - classification; Bacteria - genetics; Bias; Biodiversity; Bioinformatics; Biomedical Engineering/Biotechnology; Biomedicine; Biotechnology; Cattle; Chimerism; Ecosystems; Eukaryota - classification; Eukaryota - genetics; Eukaryotes; Gene sequencing; Genetic aspects; High-Throughput Nucleotide Sequencing; letter; Life Sciences; Metagenome - genetics; Methods; Microbial colonies; Microorganisms; Phylogeny; Polyadenine; Protozoa; Reverse transcription; Ribonucleic acid; RNA; RNA sequencing; RNA, Ribosomal, 16S - classification; RNA, Ribosomal, 16S - genetics; RNA, Ribosomal, 18S - genetics; rRNA 16S; rRNA 18S
• DOI: 10.1038/nbt.4045
• ISSN: 1087-0156; 1546-1696

More of the diversity present in any microbiome is revealed by a method to sequence 16S rRNA that avoids primer bias. Small subunit ribosomal RNA (SSU rRNA) genes, 16S in bacteria and 18S in eukaryotes, have been the standard phylogenetic markers used to characterize microbial diversity and evolution for decades. However, the reference databases of full-length SSU rRNA gene sequences are skewed to well-studied ecosystems and subject to primer bias and chimerism, which results in an incomplete view of the diversity present in a sample. We combine poly(A)-tailing and reverse transcription of SSU rRNA molecules with synthetic long-read sequencing to generate high-quality, full-length SSU rRNA sequences, without primer bias, at high throughput. We apply our approach to samples from seven different ecosystems and obtain more than a million SSU rRNA sequences from all domains of life, with an estimated raw error rate of 0.17%. We observe a large proportion of novel diversity, including several deeply branching phylum-level lineages putatively related to the Asgard Archaea. Our approach will enable expansion of the SSU rRNA reference databases by orders of magnitude, and contribute to a comprehensive census of the tree of life.