microbiology 21 22 23 24 25 26 2 | M o n i r u z z a m a n e t a l .Abstract 27Metatranscriptomics has emerged as a tool in microbial ecology that can resolve the functional 28 landscape of both prokaryotes and eukaryotes within a community. In this study, we extend the 29 potential of metatranscriptomics to probe active virus infections and virus-host relationships in 30 marine systems. Polyadenylation-selected RNA-seq data were examined from microbial 31 communities in two productive marine environments: a brown tide bloom event dominated by 32Aureococcus anophagefferens in Quantuck Bay, NY, and a diatom-dominated plankton 33 community in Narragansett Bay, RI. Active infections by diverse giant viruses (NCLDVs) of 34 algal and non-algal hosts were found at both sites. Ongoing infections of A. anophagefferens by 35 a known Mimiviridae (AaV) were observed during both the peak and decline of the bloom. 36Bloom decline was also accompanied by increased activity for viruses other than AaV, including 37 (+) ssRNA viruses. In Narragansett Bay, increased temporal reso′lution revealed active NCLDVs 38 with both 'boom-and-bust' as well as 'steady-state infection'-like ecologies. Statistical co-39 occurrence examinations of the dsDNA, ssRNA and dsRNA markers within the data revealed a 40 broad spectrum of statistically strong and significant virus-host relationships that included both 41 known as well as novel interactions. Our approach offers a method for screening the diversity 42 and dynamics of active viral infections in natural systems and develops links between viruses 43 and their potential hosts in situ. 44 45 3 | M o n i r u z z a m a n e t a l .
Significance 46Viruses are important partners in ecosystem scale ecology, yet their study to date is primarily 47 limited to single virus-host infection models in the laboratory or limited to "potential-actions" 48 derived from metagenomics analyses. Using metatranscriptomic sequences from polyadenylated-49 RNA selected samples, we have simultaneously captured information regarding eukaryotic 50 diversity and active infection by viruses with dsDNA genomes, resulting in a statistical 51 opportunity to predict "who is infecting whom". This approach further provides concurrent 52 insight regarding viruses with ssRNA and dsRNA genomes, capturing dynamics for the 53 communities of viruses infecting single-celled eukaryotes. Given the central role of these 54 plankton in global scale processes, our efforts result in a transformational step-forward regarding 55 the study of in situ virus-host interactions. 56 57