The tale of a microbial community and its infective agents: Microbial and viral interactions fuel carbon and nitrogen cycling in the hyporheic zone of the Columbia River

Rivers are increasingly recognized as important contributors to inland carbon and nitrogen budgets. While riverine microorganisms are important catalysts in these systems, studies assigning specific taxa to biogeochemical processes are limited, and particularly absent from larger order rivers. To address this knowledge gap, metagenome-resolved proteomics was performed on 43 sediment samples collected from the Columbia River (WA). We built a Columbia river Metagenome-Assembled Genome (MAG) database which included members from 12 distinct bacterial and archaeal phyla (n=102 MAGs), as well as 47 viral genera (n=171 vMAGs). Peptide recruitment to unique MAGs demonstrated that organic nitrogen (N) degradation (proteolysis) was active and conserved across 6 sites and 5 depths, and likely generated amino acids and ammonium evident in our metabolite and geochemical data. Ammonium is required to sustain proteomically-verified chemolithotrophic nitrification by ammonia-oxidizing Crenarchaeota and nitrite-oxidizing Nitrospirota. This coupled N mineralization-nitrification may represent sources of oxidized nitrogen that members of the Gammaproteobacteria incompletely denitrify, contributing to nitrous oxide fluxes measured from these sediments that range from -772.3 to 1235.9 nmol m ² h ^-1. Demonstrating the importance of these N-cycling microorganisms to system biogeochemistry, the relative abundance of Nitrospirota genomes, as well as their hypothesized viral predators, predicted the abundance of total N in these sediments. We also showed that the reconstructed viral genomes contained genes that extended the carbohydrate substrate range for their hosts, offering a metabolic advantage in these organic carbon depleted sediments. Spatiotemporal sampling of Columbia River sediments over 3 years and a lateral distance of 1 mile revealed broad conservation of these microbial and viral genomes, suggesting stable biogeochemical roles for these community members in fluvial ecosystems. In summary, this meta-proteomics approach uncovered a coupled N mineralization-ammonification that could generate cryptic nitrate sources not-yet accounted for in reactive transport models, and defined new roles for viral predation and gene content as modulators of river sediment biogeochemistry.