DNA-templated assembly of the bacterial flagellar motor's cytoplasmic ring

The bacterial flagellar motor is one of the most complex protein machines found in nature and how it self-assembles and produces force are very much open questions. In this thesis, I study the constituent parts of the motor in vitro, with the focus being primarily on FliG, a protein which helps to form the motorâs cytoplasmic ring. The experimental approach employs a DNA scaffold to direct the in vitro formation of the FliG ring, with the aim of determining both the number of proteins in the complex and the means by which they associate to form a circular structure. One leading theory states that the FliG proteins interact through a process of domain-swap oligomerization and the primary goal of this work is to search for evidence of this process with the aid of in vitro DNA templates. To attach the FliG protein to the underlying DNA structures, I produced FliG-DNA conjugates through two different means. First, I bound both tris- and pentakis-NTA-modified DNA strands to the Histidine-tag of FliG and showed that these conjugates could be arranged on simple, linear DNA templates. In order for the proteins to be arranged with the spacing observed in the cytoplasmic ring however, a different conjugation strategy was required in which a maleimide-modified DNA strand was reacted with a single-cysteine FliG mutant. After issues relating to Histidine-tag-mediated FliG dimerization were resolved, these conjugates were able to be organized on both linear DNA strands and origami tiles with the desired spacing between the proteins. No interactions between either template-bound or free-floating FliG proteins could be observed however, meaning a new strategy was needed. One model of cytoplasmic ring assembly predicts that binding of FliG to the C-terminal domain of the membrane-bound FliF protein induces a conformational change in both partners which then triggers interactions between neighbouring FliG proteins. To test this, I designed a peptide corresponding to the FliFC domain of Salmonella and E. coli FliF and incorporated it into my system. Several biophysical assays indicated that the FliFC peptide stably bound to purified wild-type Salmonella FliG and that binding to the peptide increased the overall stability of the protein. This constitutes the first confirmation that the cytoplasmic domain of FliF is sufficient for FliG binding in enteric bacteria. Following this discovery, a maleimide-modified DNA strand was then conjugated to a single-cysteine version of the peptide and the peptide-DNA conjugates organized with various stoichiometries and spacing on DNA templates to create an array of âbinding platformsâ for the FliG protein. FliG bound to the organized peptides in a one-to-one fashion but interactions between the organized FliG proteins were still not detected. In a final effort to stimulate FliG-FliG binding, a radially-symmetric DNA nanostructure was designed with dimensions and a geometry which matched those found in the cytoplasmic rings of bacteria. Gels and AFM microscopy proved that the 12 designed strands assembled correctly into the desired structure and that FliFC-DNA conjugates with FliG proteins bound could be organized on the circular arrangement of binding sites around the structureâs outer rim. Even in such an arrangement, interactions between neighbouring proteins could not be observed, and though higher concentrations of FliFC and FliG were tested, issues with purification prevented accurate characterization of these complexes. Thus, though the domain-swap model for FliG oligomerization was therefore neither confirmed or denied, the use of a DNA scaffold to organize protein molecules and study their interactions holds promise as a technique to investigate the proteins of other biological systems and will no doubt be employed to great effect in the coming years.