Inhibition and Allosteric Signaling of Transcription Activation by Bacterial AraC Family Activator Proteins

Abstract

AraC family transcriptional activators are defined by a 100-amino acid DNA-binding domain (DBD) containing two DNA binding helix-turn-helix (HTH) motifs. My research focused on three AraC family proteins: RhaR (activator of the E. coli L-rhamnose catabolic regulon), VirF (activator of expression of the Shigella flexneri type three secretion system), and ToxT (activator of Vibrio cholera virulence gene). By using fluorescence-based thermal shift assay, and intrinsic tryptophan fluorescence assay, I have shown that small molecule inhibitor SE-1 directly binds to VirF and RhaS-DBD. Mutagenesis studies of residues in RhaS support the prediction of docking that SE-1 is likely bound to a small pocket between the two HTH motifs. With a final goal to provide direct evidence of the position on the ToxT protein that serves as the SE-1 binding site, ToxT was crystalized under solution conditions different than the previously published ones, and the structure of ToxT was determined to a higher resolution than the published structure. A region that was missing from the previously determined structure now can be traced entirely. Using Electrophoretic Mobility Shift Assay (EMSA), SE-1 was found to inhibit DNA binding by ToxT. Co-crystallization trials of SE-1 and ToxT were carried out. Unfortunately, I was not able to obtain any crystals of ToxT-SE-1. In collaboration with Dr. Jeff Aubé, I have tested SE-1 analogues with the ultimate goal of optimizing the potency and specificity of SE-1. Unfortunately, no analogs were found with increased VirF inhibition potency relative to SE-1. In the process of synthesizing analogs, we found that SE-1 and its analogues converted to corresponding quaternary salts in aqueous solution, and the quinolinium salt was responsible for the observed inhibition by SE-1. In this study, I also investigated the mechanism by which RhaR responds to L-rhamnose. During these studies, I discovered that the RhaR start codon was previously annotated 30 codons upstream of the true start codon. The equilibrium binding affinity of RhaR to its full and half DNA binding sites, of the isolated RhaR-DBD to the half-site, and the rhamnose dependence of DNA bending by RhaR were measured. The findings support a model in which RhaR NTD increases the DNA binding affinity of each RhaR protomer DBD, independent of rhamnose; and that rhamnose signalling primarily increases the positive cooperativity of DNA binding by the two DBDs in a RhaR dimer. These findings suggest a model for the mechanism of allosteric rhamnose signalling in RhaR. To provide more structural information about RhaR protein, the structures of RhaR-NTD were determined in the presence and absence of L-rhamnose. The 2.05Å rhamnose-bound RhaR-NTD structure showed that the protein forms an antiparallel dimer, and shared a fold that was similar to the AraC-NTD, binding its respective sugar L-rhamnose within a β-barrel. In addition, a Ni2+ ion, which has not been seen in other AraC family protein structures, was present in the sugar-binding, cupin superfamily, motif of RhaR. A rhamnose-free structure was also solved to 1.73 Å, and in this structure, a loop region that is involved in rhamnose binding was completely disordered. A second loop region also has minor structural changes. Each of the two regions with rhamnose-dependent structural changes is predicted to be at the interface between the RhaR NTD and DNA-binding domain, suggesting their potential involvement in rhamnose allosteric signaling. No differences were observed in the RhaR N-terminal arm region in the fully and partially rhamnose-occupied structures, suggesting that RhaR rhamnose-dependent allosteric signaling shares some features with the ‘light switch’ model of AraC, but differs in other features.