Viral Community Dynamics and Functional Specialization in the Pacific Ocean

Abstract

Viruses are the most abundant biological entity on Earth and outnumber their hosts ten-to-one. Ocean viruses (phages) impact bacterial-driven global biogeochemical cycles through lysis, manipulating host metabolism, and horizontal gene transfer. However, knowledge of virus-host interactions and viral roles in ecosystems remains limited due to few cultured marine phage genomes and non-quantitative culture-independent metagenomes. Here, I develop and apply novel and well-tested bioinformatic techniques to explore Pacific Ocean viral communities using quantitative datasets derived from rigorously-tested preparation methods. To evaluate concentration and purification methods, I examined triplicate metagenomes from a single ocean sample using four protocols. Concentration protocols showed statistical differences in taxonomy whereas purification protocols did not. Specifically, TFF-concentrated metagenomes contained trace bacterial contamination and had fewer abundant taxa as compared to FeCl₃-precipitated metagenomes. K-mer analysis using the complete dataset revealed polymerase choice defined access to "rare" sequences.To explore unknown viral sequences, I organized known and unknown sequence space into 27K high-confidence protein clusters (PCs) from 32 diverse Pacific Ocean Virus (POV) metagenomes, which doubled available PCs and included the first pelagic deep-sea viral metagenomes. Using PCs as a whole-viral-community diversity metric revealed decreases from coastal to open ocean, winter to summer, and deep to surface, that correlate with data from microbial genetic diversity markers (no parallel viral markers exist).Biologically, POV metagenomes showed that viruses likely reprogram central metabolic pathways in microbial communities far beyond the "photosynthesis viruses" paradigm. Gene distribution patterns from 35 viral gene families (31 new) revealed niche-specific (photic vs aphotic zone) altered pathway carbon flux presumably optimized to best locally generate energy and drive viral replication. Further, these PCs define the first "core" (180 genes) and "flexible" (423K genes total) viral community genome. Functionally, core genes again suggest niche-differentation with extensive Fe-S cluster-related genes for electron transport and metabolic enzyme catalysis in photic samples, and manipulation of host pressure-sensitive genes in aphotic samples. Taxonomically, these data deconstruct the culture-based paradigm that tailed viruses dominate in the wild - instead they appear ubiquitous, but not abundant.