Award: OCE-1820652

Overview Viruses, primarily of bacteria, launch 1027 infections in the ocean every day, making them the most pervasive and successful "predators" on the planet. These virus-host interactions transform biogeochemical cycles in the ocean through cell lysis and virally-driven expression of key metabolic genes. For uncultured microorganisms, techniques to identify virus-host interactions, such as polonies or microfluidic digital PCR, require specialized expertise and equipment, or yield few positive interactions. Other non-specific culture-independent techniques, such as proximity ligation and single-cell genomics, are exciting non-specific approaches but are costly and inefficient because the vast majority of reactions or sequence products are expected to be negative or contain no viral-host information. This can make it difficult to determine viral-induced mortality associated with specific virus-host interactions or determine how environmental factors influence these interactions when positive interactions make up a small fraction of the resulting data. To address this problem, we successfully adapted a powerful molecular tool, epicPCR, to identify virus-host interactions in uncultured microorganisms. This cost-effective and targeted technique provides sufficient detail to discover novel and important fundamental relationships between viruses, host, and the environment. This technique takes advantage of the physical proximity of the virus and host DNA to fuse viral and host marker genes using PCR within emulsion droplets to achieve single-cell specificity. We targeted viruses that were assumed to infect Cyanobacteria based on bioinformatic inference but discovered that the dominant host for these poorly characterized estuarine viruses was Actinobacteria, highlighting the need for experimental techniques to identify the host for many uncultured viruses. By applying this technique to a time-series in the Chesapeake Bay watershed, we discovered that virus-host interactions of different viral clades and their dominant host (Rhodoluna) were associated with tidally driven changes in salinity and organic material, which are environmental factors known to influence infections from culture-based and bulk community measurements but had not previously been demonstrated for uncultured virus-host pairs. Tracking virus-host interaction dynamics also revealed that multi-host viruses had significantly longer periods with observed virus-host interactions, whereas single-host viruses were observed interacting with hosts at lower minimum abundances, suggesting more efficient interactions. Significance and Broader Impact The COVID pandemic has demonstrated how important it is to have a cost-effective test of viral infections. Without widespread COVID testing, it would have been impossible to understand the magnitude and scope of the pandemic. Our work provides a cost-effective diagnostic test of virus-host interactions for the vast majority of viruses and bacteria that cannot be cultured. This test can be applied to track viral infections for large numbers of samples in any aquatic environment to reveal biologically meaningful patterns in the data. Our technique can experimentally determine the host range for uncultured viruses, which may reveal additional differences between single-host (specialist) and multi-host (generalist) viruses. The broader impacts of this award include training of one postdoctoral associate, one Master's student and two undergraduate researchers, including an underrepresented minority student. Last Modified: 10/28/2021 Submitted by: Sarah Preheim