Carbohydrate-binding proteins from marine bacteria

Marine phytoplankton is responsible for about half of the CO2 fixation on Earth. Algal photosynthesis results in the production of organic carbon, a substantial fraction of which is in the form of polysaccharides. Bacterial utilization of polysaccharides is a highly relevant process, which constitutes a major carbon and energy transfer in the ocean. Bacteroidetes, key marine polysaccharide degraders, are known to employ a complex protein system to bind, transport and digest polysaccharides. It was first characterized for the starch utilization system in Bacteroides thetaiotaomicron. In contrast to terrestrial or human gut microbes, polysaccharide degradation by marine bacteria remains largely unexplored. The main aim of my thesis was to investigate the molecular details of microbial polysaccharide utilization in the ocean. More specifically, during my Ph.D. thesis, I investigated carbohydrate-binding proteins involved in the recognition of substrates at the surface of bacterial cells. Based on the knowledge from homologous systems and bioinformatic predictions, we assumed that marine Bacteroidetes assemble outer membrane complexes composed of surface-glycan binding proteins to acquire polysaccharides. Previous experiments using fluorescently-labeled polysaccharides showed that marine Bacteroidetes take up polysaccharides in a selfish manner. These bacteria use surface-associated enzymes and binding proteins to partially degrade polysaccharides and minimize production of freely diffused hydrolysis products. For this approach, bacteria must have evolved a highly efficient and selective binding apparatus, which I studied in detail in this thesis. These in-depth analyses were necessary to evaluate the potential of carbohydrate-binding proteins as novel glycan probes. In the first manuscript, I present the characterization of GMSusD protein. We focused our analyses on proteins putatively specific to the highly abundant marine polysaccharide - laminarin. Biochemical and structural analyses on the GMSusD from Gramella sp. MAR 2010 102 revealed the predicted laminarin binding. Surprisingly, the protein was specific to a particular type of laminarin structure. There is a big discrepancy between vast a omicsa sequence data and functionally or structurally characterized proteins. To provide more accurate support for bioinformatic predictions we performed structure-guided alignment of metagenomes of global surface water datasets using the structure of GMSusD as a guide. We found SusD-like proteins with structurally conserved residues of the binding site in different locations in the ocean, suggesting a similar manner of laminarin recognition by these proteins. In the second manuscript, we identified two additional laminarin-binding proteins from the same planktonic bacteria. Upstream to the GMSusD, in the gene cluster called a Polysaccharide utilization locus, there were two genes encoded with previously unknown VI function. After excluding their enzymatic activity, we analyzed binding abilities. Based on the identified laminarin binding activity and predicted three-dimensional structures, we propose that these two proteins, GMSusE and GMSusF, belong to a highly unexplored group of SusEF- like binding proteins. Finally, I applied medium throughput expression of recombinant putative carbohydrate-binding proteins to investigate their potential as glycan probes. The field of marine glycobiology needs to be extended, since little is known about structures of polysaccharides present in the ocean. Thus, we proposed taking advantage of bacterial proteins, which are expressed in response to algal blooms in the North Sea. We investigated a library of forty-seven constructs resulting in the production of twelve soluble recombinant proteins, the binding of which was tested with environmental algal extracts and well-defined polysaccharide controls. These analyses allowed us to discover four novel carbohydrate- binding proteins specific to laminarin, a-mannan and b-mannan. However, we encountered some limitations of this approach, which are discussed in the third manuscript. The research performed in this thesis contributes to our greater quest to understand algal carbohydrate binding by marine microbes, which is a crucial mechanism for bacterial polysaccharide utilization and therefore key in the marine carbon cycle