CRISPR spacers reveal diverse and abundant Thermococcales viruses in hydrothermal vents.

Viruses are the most pervasive biological entities on Earth and they profoundly shape host ecology and evolution. However, for many microbial lineages, knowledge of their viromes remains limited, especially for those inhabiting remote environments, including deep-sea ecosystems. Here, we leverage one of the most extensively cultivated and genomically characterized archaeal lineages, the Thermococcales, to identify novel viral genomes. By utilizing CRISPR spacers from isolates and spacer arrays reconstructed from metagenomes, we mined mobile genetic elements (MGEs) in 1,172 publicly available and newly sequenced hydrothermal vent metagenomic datasets. Comparative genomics and identification of viral hallmark proteins revealed 620 viral genomes across 19 taxonomic families, most of which were previously undescribed. Structural modeling of major capsid proteins (MCPs) revealed diverse virion morphotypes, including viruses with spindle-shaped, head-tailed, icosahedral, filamentous, ovoid and bacilliform virions, greatly expanding the previously limited Thermococcales virome. Family-level comparisons uncovered extensive flux of virus-encoded replication proteins that are evolutionarily and structurally distinct from host homologs, as well as dramatic variation in glycan-binding lectins suggestive of diverse infection strategies. Together, our results substantially broaden the Thermococcales virosphere and demonstrate the power of combining cultivated isolates with culture-independent, CRISPR-guided metagenomics to interrogate archaeal virus diversity and evolution.