Jozef Nissimov, University of Waterloo

Isaac Meza-Padilla, Cody C. Collis, Victoria Lee, Brendan J. McConkey, Andrew C. Doxey, Jozef I. Nissimov

Discovering & Evolving

Viruses control the abundance and diversity of other microbes in aquatic habitats. Virus infection of certain microbes may also have unexpected effects. We recently showed that cyanophages that infect the freshwater bloom-forming toxic cyanobacterium Microcystis aeruginosa, contrib-ute to the release of toxins upon cell lysis. However, our understanding of how viruses achieve this is still limited; often, >80% of genes encoded by these viruses are cryptic (i.e., code for hy-pothetical or membrane proteins with an unknown function), having little to no sequence-based homology in public databases. This includes sequenced genomes of isolates, and those assem-bled de novo through metagenomics. So, to fully understand viral impacts on aquatic systems, we need to know more about their genetic repertoire. Here, we employed AlphaFold and Fold-seek, to gain novel information on the functional role of cryptic proteins of two cyanophages- Ma-LMM01 and CrV-01T. Models of cryptic virus proteins were matched to novel putative functions, including photosystem I, viral anti-defense, and a cyanobacteriochrome, a photoreceptor protein normally found in cyanobacteria. We also show that the non-bleaching A (NblA) protein encoded by the Ma-LMM01 phage is predicted to have a significantly higher binding affinity for host phy-cocyanin (αβ)6 hexamers, compared to the host homolog, which may represent a selective ad-vantage for the virus, whose infection cycle requires an increased phycobilisome degradation rate that is likely not fulfilled by the NblA of the host. These insights highlight the importance of structure-based homology in identifying novel functions in viruses to better understanding their ecological and evolutionary role.