Ecogenomics of groundwater viruses suggests niche differentiation linked to specific environmental tolerance

Viruses are ubiquitous microbiome components, shaping ecosystems via strain-specific predation, horizontal gene transfer and redistribution of nutrients through host lysis. Viral impacts are important in groundwater ecosystems, where microbes drive many nutrient fluxes and metabolic processes, however little is known about the diversity of viruses in these environments. We analyzed four groundwater plasmidomes and identified 200 viral sequences, which clustered into 41 ~ genus-level viral clusters (equivalent to viral genera) including 9 known and 32 putative new genera. We use publicly available bacterial whole genome sequences (WGS) and WGS from 261 bacterial isolates from this groundwater environment to identify potential viral hosts. We linked 76 of the 200 viral sequences to a range of bacterial phyla, the majority associated with Proteobacteria, followed by Firmicutes, Bacteroidetes and Actinobacteria. The publicly available microbial genome sequences enabled mapping bacterial hosts to a breadth of viral sequences. The WGS of groundwater isolates increased depth of host prediction by allowing identification of hosts at the strain level. The latter included 4 viruses that were almost entirely (>99% query coverage, >99% identity) identified as integrated in the genomes of specific and strains, resulting in very high-confidence host assignments. Lastly, 21 of these viruses encoded putative auxiliary metabolite genes for metal and antibiotic resistance, which might drive their infection cycles and/or provide selective advantage to infected hosts. Exploring the groundwater virome provides a necessary foundation for integration of viruses into ecosystem models where they act as key players in microbial adaption to environmental stress.