Using metaproteomes and models to quantify cellular trade-offs in phytoplankton

Abstract

Phytoplankton fuel biogeochemical processes in the ocean and are key players influencing the global climate. All biogeochemical processes mediated by microbes are ultimately underpinned by gene expression. In this thesis, I aim to connect gene expression to biogeochemically important cellular processes in Southern Ocean phytoplankton. First, I identify and model pervasive biases in metaproteomic analyses and develop methods for overcoming them, leading to more robust inferences (Chapter 2 and Chapter 3). In Chapter 2, I develop a computational model for predicting cofragmentation bias and use this model to study how cofragmentation impacts inferences in metaproteomics. In Chapter 3, I delineate ‘proteomic traits’ across microbial taxa in an Antarctic phytoplankton bloom, connecting differences in gene expression patterns to ecological strategies. I also highlight the importance of database choice and quantify its implications for metaproteomic conclusions. In Chapter 4, I develop a proteomic allocation model to quantify trade-offs associated with iron and manganese bioavailability, and reframe micronutrient-controlled growth in the ocean as a function of cellular costs and constraints. This model offers a novel framework for leveraging metaproteomic data to learn about cellular processes in phytoplankton and for inferring taxon-specific rates and biogeochemical metrics. A key unknown in this model is the various antioxidant systems used by phytoplankton. I, therefore, review various antioxidant mechanisms and synthesize their contributions to cellular elemental stoichiometry in phytoplankton (Chapter 5). Finally, in Chapter 6, I use metaproteomics to determine environmental controls on ribosomal mass fraction across two taxonomic groups in the Amundsen Sea Polynya.