UC Berkeley

This thesis addresses a diversity of questions regarding the structural details of sigma54 transcriptional activation, and the function of sigma54 activation in the hyperthermophile Aquifex aeolicus. In order to place each topic in its appropriate context, a general introduction is provided in the first chapter, and supplemented with additional, more detailed introductions in each subsequent chapter. The second chapter reflects the central project of this thesis, the determination of the structure of the DNA-binding domain of an NtrC-like sigma54 transcriptional activator protein in Aquifex aeolicus, in complex with its high-affinity DNA binding site. Although this project was attempted by NMR, its structure was ultimately solved by X-ray crystallography. This structure, which shows slight DNA-bending, is compared to the recent structure of the homologous Fis protein in complex with DNA.

In the third chapter, I describe a cross-comparison of DNA-binding domains from Aquifex aeolicus, including new structures of the DNA-binding domains of NtrC1 and NtrC2, and comparisons to NtrC4, ZraR and NtrC from Salmonella enterica serovar typhimurium, and Fis from Escherichia coli. These structures of DNA-binding domains from a single family enable a detailed comparison of structural changes that tune protein function. A trend is noted, in which larger dimerization interfaces in the DNA-binding domains appear to correlate with smaller dimerization interfaces in the other domains of the protein. The additional structure of the full-length NtrC1 protein is presented, but the DNA-binding domains were missing from the density, providing evidence for the flexibility of the linker between the central and DNA-binding domains. Together with structural information about the CD linker regions at the N-termini of the DNA-binding domains, I conclude that the CD linker regions are generally unstructured in the inactive state, functioning as tethers to bring the RC domains of NtrC into appropriate local concentrations for activity.

In the fourth chapter, I depart from sigma54 transcriptional activators, and discuss NMR studies on the sigma54 factor. The Darst lab at Rockefeller University has produced a crystal structure of the full-length intact sigma54 factor in complex with DNA. However, poor data quality prevents regions of the molecule from being traced unambiguously through the density. By applying new methods in NMR including TROSY spectroscopy and specific isotopic labeling, I attempted to resolve small regions of structure in this ambiguous region. However, these studies were complicated by protein aggregation.

The fifth and final chapter takes a step back from structural perspective and compiles and discusses our current understanding of sigma54 regulatory networks in Aquifex aeolicus. This analysis of genes with bioinformatics and biochemical techniques led me to discover that NtrC3, an unstudied sigma54 transcriptional activator, binds upstream of the dhsU gene, which is responsible for sulfur metabolism. NtrC3 also appears to be associated with a heme-binding, soluble histidine kinase, HksP4. In order to explore this two-component regulatory system further, the HksP4 protein was prepared and its gas-binding properties were studied in collaboration with Dr. Brian Smith of the Marletta lab. We find that the heme binds O2, NO, and CO gases in the reduced ferrous iron state, but did not observe any autophosphorylation activity in any state. I conclude that an additional element, such as an additional activator, or more physiological conditions, would be necessary for activity.