Unlike prokaryotes, which are infected almost exclusively by DNA viruses, eukaryotes are susceptible to both DNA and RNA viruses, with the latter comprising the majority. This phenomenon has been attributed to the emergence of the nucleus, which imposes an additional barrier to DNA viruses, and to the expansion of cellular membranes, which create a favorable environment for RNA viruses. In this scenario, early eukaryotes may have evolved to use their RNase III arsenal to engage incoming RNA viruses and cleave hairpins formed during their replicative process. This mechanism alone would have provided some positive selection, but the more significant advancement would later come with the capacity to couple these RNA cleavage products to an otherwise nonspecific nuclease, such as the Argonaute family members—a development representing the earliest form of RNAi. In fact, based on homology, it appears that antiviral RNAi arose before the divergence of animals and plants, with the underlying biology later independently co-opted for developmental purposes in the form of miRNAs. It is thus interesting that vertebrates encode two RNase III nucleases, Dicer and Drosha, with the latter exhibiting greater homology to the ancestral family and uniquely demonstrating antiviral activity. This hypothesis not only provides a framework for how antiviral pathways may have formed over evolutionary time, but it also explains why RNase III nucleases consistently demonstrate antiviral activity across all domains of life.

CRISPR-Cas systems serve as key antiviral defense mechanisms in bacteria and archaea. In an effort to mine for additional defense systems in archaea, we analyzed the transcriptomics of virally infected Methanococcus maripaludis. Bulk RNA sequencing of M. maripaludis infected with a novel member of the myovirus family revealed ~250 genes that were significantly upregulated. Focusing on DNA- and RNA-binding proteins that were upregulated in response to infection, we synthesized 30 genes as codon-optimized open reading frames for expression in mammalian cells. These genes were tested against seven distinct virus challenge models to ascertain antiviral activity, revealing a subset that could either enhance or inhibit virus replication:

• Bulge-Helix-Bulge (BHB) protein, a product of Operon 585-1, showed selective inhibition of +ssRNA viruses (e.g. poliovirus, Langat virus). This gene product is a member of the dsRNA-specific nucleases related to the RNase III family, which has been implicated in both the Type II CRISPR-Cas system and in the processing of RNAi. Motivated by this finding, we cloned members of this gene family from representative kingdoms to assess whether the antiviral activity was conserved. Our results demonstrated that an inherent property of these proteins was to engage and induce the cleavage of RNA hairpins commonly found in the genomes of +ssRNA viruses. Together, these data may reflect the evolutionary history of antiviral systems, suggesting that this nuclease family may have been repurposed during eukaryogenesis.

• Operon 169 (O160), a putative RNA binding protein that was induced >100-fold in M. maripaludis, was found to potently block influenza A virus (IAV) infection when introduced into mammalian cells. The closest orthologue of this gene in humans is the ribosomal protein, RPL30. In a manner similar to how we approached RNase III family members, we synthesized codon-optimized components of the RPL30 family of ribosomal proteins, which are conserved in all three domains of life, to assess whether expression could disrupt IAV replication. Members of this family from Plasmodium to Xenopus were expressed in human lung alveolar cells and were infected with the 2004 pandemic H1N1 IAV. These data revealed that only archaeal homologs induced viral inhibition, a phenotype we later demonstrated was due to their ability to block production of NS2, an alternatively spliced viral product required for replication. While O169 effectively inhibited every strain of IAV we tried (n=8), it had no impact on host splicing or the mammalian cell itself, suggesting it could represent a universal inhibitor of IAV infection.