Structure-Function Investigation of Proteins Involved in Cellulose Biosynthesis by Escherichia coli

Bacteria thrive within multicellular communities called biofilms consisting of a self-produced matrix. Biofilm matrices improve bacterial adherence to surfaces while creating a barrier from host immune responses, disinfectants, antibiotics and other environmental factors. Persistent colonization by the widely distributed pathogens, Escherichia coli and Salmonella spp., has been linked to production of biofilms composed of the exopolysaccharide cellulose. Cellulose-containing biofilms are also important to Acetobacter, Sarcina, Rhizobium and Agrobacterium species to form symbiotic and pathogenic interactions. In Enterobacteriaceae, two operons (bcsABZC and bcsEFG) are proposed to encode for proteins that form a cellulose biosynthetic complex that spans the bacterial cell wall. Using established recombinant DNA techniques, crystallography and functional assays, the overarching objectives of this research included the investigation of the structures and functions of the BcsE and BcsG proteins to gain insight into how they contribute to the Bcs system. The cytoplasmic protein BcsE has been shown via knockout studies to be required for optimal cellulose biosynthesis and recently the C-terminus was proven to be the second protein domain, after PilZ, dedicated to c-di-GMP binding in Enterobacteriaceae. The N-terminal structure and function of BcsE are still uncharacterized. One hypothesis for the function of BcsG is that it is involved in the labelling of cellulose with phosphoethanolamine (PEA) during export due to its homology to other characterized proteins. For example, the external modification with PEA is a strategy that allows organisms like Neisseria gonorrheae to evade components of the host immune response. However, the structure, cellular localization and specific mechanism of action of BcsG are yet unknown. To gain insight into the hypothetical properties of BcsE and BcsG, bioinformatics analyses were first conducted. The following research focused on the structure-function characterization of these proteins using recombinant truncated constructs for hypothetical N- and C-terminal domains. While practical quantities of BcsE constructs could be expressed and purified, these constructs proved challenging to isolate in sufficient purity and concentration for structural analyses. High yields of the C-terminal, soluble BcsG construct (amino acids 164 – 559) were ideal for structural and functional analyses. Malachite green-based colorimetric phosphate detection assays supported the bioinformatics analyses that the soluble C-terminus of BcsG has phosphatase activity in the presence of ATP, GTP, CTP and PEA. Metal dependency and pH tests showed that optimal BcsG activity occurs at pH 7.5 with a magnesium cofactor (2.03 x 10-1 +/- 0.008 nmol/mg/min) which supports bioinformatics predictions. Using a BcsG1-559-GFP hybrid, the localization of the soluble C-terminus of BcsG was shown to reside in the periplasm of E. coli. This localization aligns with bioinformatics analysess and would give BcsG a logical vantage point for cellulose modification during export from the cell. Numerous crystallization screens were attempted for BcsE and BcsG constructs. High quality BcsG164-559 native protein crystals were achieved with resolutions as sharp as 2.1Å as measured by X-ray analysis at the Canadian Light Source. Experimental phasing with heavy metal soaking and selenomethionine labeling techniques were attempted in search of missing phase information for BcsG164-559. These techniques have shown promise; however, experiments are ongoing. Future studies with BcsG should continue phasing experiments, test more substrates from the PEA metabolism pathway and attempt active site characterization. Future BcsE research should focus on N-terminal functional investigations and structural experiments for the N- and C-termini