Identification of key virus–prokaryote infection pairs that contribute to viral shunt in a freshwater lake

Abstract

Viruses infect and kill productive prokaryotes in a density- or frequency-dependent manner and affect carbon cycling. However, the effects of the stratification transition, including the stratified and de-stratified periods, on the changes in prokaryotic/viral communities and the interactions among them remain unclear. We conducted a monthly survey of the surface and deep layers of a large and deep freshwater lake (Lake Biwa, Japan) for a year and analyzed the prokaryotic production and prokaryotic/viral metagenome. Our analysis (including 1 608 prokaryotes and 13 761 viruses) revealed that 19 prokaryotic species, accounting for ~40% of total abundance, might be suppressed by viruses when prokaryotic production is higher. This suggests that a small proportion of prokaryotes contribute to a large amount of prokaryotic abundance, and these prokaryotes are infected and lysed by viruses, driving the viral shunt in the freshwater lake. Furthermore, we found that annual vertical mixing might yield a similar rate of community change between the surface and deep layers. This finding might be valuable in understanding how the communities change when the stratification of freshwater lakes is affected by global warming in the future.

Contrasting defense strategies of oligotrophs and copiotrophs revealed by single-cell-resolved virus–host pairing of freshwater bacteria

“…The first (Lake Biwa viral contigs, LBVC) was reconstructed from cellular (0.2-5 μm size fraction) and virion (< 0.2 μm) metagenomes collected monthly from June 2016 to February 2017 from the epilimnion and hypolimnion (6). The second (Lake Biwa viruses, LBV) was reconstructed from virion metagenomes collected monthly from September 2018 to December 2019 from the epilimnion and hypolimnion (8). Contigs > 10 kb were selected from both datasets and quality-controlled using CheckV, as described above.…”

“…We mapped reads of published metagenomes to the viral contigs to evaluate their spatiotemporal distribution in the lake. We used two datasets: virion (< 0.2 μm) metagenomes collected monthly from December 2018 to November 2019 from the epilimnion and hypolimnion (8), and cellular (0.2-5 μm) metagenomes collected monthly from May 2018 to April 2019 (7). For metagenomic read mapping, the CoverM option "--min-readpercent-identity 92" was used.…”

“…An important achievement of this study is the identification of novel virus-host pairs by the detection of viral signals in SAGs, which is unprecedented in a freshwater lake. Although large-scale viral metagenomic surveys in freshwater lakes routinely recover thousands of viral genomes (6,8,10,50), the hosts of most of these viruses are unknown, which limits understanding of their ecology. Furthermore, host prediction relies mainly on gene phylogeny, which does not resolve the host to a low taxonomic level.…”

“…To overcome the limitations of metagenomics, we conducted single-cell genomic analyses of environmental bacterial assemblages. We targeted bacterioplankton and their viral community in Lake Biwa, Japan, where metagenomic studies have reconstructed > 500 bacterial and > 7000 viral genomes and characterized their spatiotemporal distributions (6)(7)(8)(9). Although these studies identified quantitatively significant bacteria and viruses in the lake, the limited resolution of metagenomics hindered the prediction of virus-host pairs and the elucidation of their interactions.…”

“…This approach has been used to identify novel virus-host pairs in marine (12)(13)(14) and hot spring (15,16) systems but not in freshwater lakes. Based on published metagenomic data (7,8) and high-quality SAGs generated from multiple seasons and depths using state-of-the-art gel bead-based SAG (SAGgel) technology (17), we explored virus-host interactions in the lake at single-cell-resolution. The results identified many novel virus-host pairs, including those involving the most quantitatively significant freshwater bacterioplankton lineages, which were previously overlooked in metagenomic studies.…”