Long-read shotgun metagenome sequencing using PromethION uncovers novel bacteriophages, their abundance, and interaction with host bacterial immunity in the oral microbiota

Abstract

Bacteriophages (phages), or bacterial viruses, are very diverse and highly abundant worldwide, including human microbiomes. Although a few metagenomic studies have focused on oral phages, they relied on short-read sequencing. Here, we conducted a long-read metagenomic study of human saliva for the first time using PromethION that requires a smaller amount of DNA than PacBio. Our analyses, which integrated both PromethION and HiSeq data of >30 Gb per sample, revealed N50 ranging from 187-345 kb and thousands of contigs with >1 kb accounting for > 99% of all contigs on which 94-96% of HiSeq reads were mapped. We identified hundreds of viral contigs (95 phages and 333 prophages on an average per sample); 0-43.8% and 12.5-56.3% of the "most confident" phages and prophages, respectively, didnt cluster with those reported previously and were identified as novel. Our integrated analyses identified highly abundant oral phages/prophages, including a novel Streptococcus phage cluster and nine jumbo phages/prophages. Interestingly, 86% of the phage cluster and 67% of the jumbo phages/prophages contained remote homologs of antimicrobial resistance genes, suggesting their potential role as a source of recombination to generate new resistance genes. Pan-genome analysis of the phages/prophages revealed remarkable diversity, identifying 0.3% and 86.4% of the genes as core and singletons, respectively. Functional annotation revealed that the highest fraction of the core genes was enriched in phage morphogenesis, followed by the fraction enriched in host cellular processes. Furthermore, our study suggested that oral phages present in human saliva are under selective pressure for escaping CRISPR immunity. ImportanceDespite the abundance and grave implications oral bacterial viruses in health and disease, little is known regarding the different groups of oral bacterial viruses, their relative abundances under various conditions, and their activities. We provided answers to these questions for the first time utilizing a recently developed sequencer that can capture and sequence long DNA fragments, including viruses, and requires only a small amount of DNA input, making it suitable for analyzing human oral samples. We identified hundreds of viral sequences, including "jumbo" viruses and a distinctive group of highly abundant oral viruses, which often contained parts of antimicrobial resistance genes; the entire repertoire of these viral genes showed remarkable diversity and supported a recently proposed hypothesis that phages modulate oral microbiota through multiple mechanisms. We also revealed genomic signs of coevolution of viruses and host bacteria that have been missed in large viromic studies in humans.