Haematopoiesis is the process of blood cell formation from self-renewing haematopoietic stem cells (HSCs) during embryonic development, and their replenishment throughout adulthood. HSC lineage commitment and cellular differentiation is tightly orchestrated by cell-signalling pathways and complex transcription factor (TF) networks. Central to transcriptional regulation of haematopoiesis is the master TF PU.1 which directs myeloid commitment and macrophage differentiation via the upregulation of myeloid-specific genes, including the macrophage colony stimulating factor receptor (M-CSFR). PU.1-mediated regulation of the M-CSFR promoter is influenced by protein-protein interactions with other TFs, including RUNX1, c-Jun, E47 and GATA1. Aberrations of these interactions are implicated in acute myeloid leukaemia and other pathologies. Whilst the interactions between PU.1 and these other TFs have been mapped to their respective DNA binding domains (DBDs), no structural characterisation of these complexes exists. The work in this thesis focuses on the structural and functional characterisation of these protein-protein and protein-DNA interactions in the context of M-CSFR promoter regulation. The ability of the implicated DBDs to form complexes was characterised using size exclusion chromatography and isothermal titration calorimetry. Contrasting previous studies, our work strongly suggests the DBDs alone are unable to form complexes at biologically relevant affinities, implying other factors may be important for mediating these interactions. ab initio modelling from small-angle X-ray scattering revealed the DBDs of RUNX1 and PU.1 form a complex on the M-CSFR promoter, mediated by protein-DNA but not protein-protein interactions. Our work also demonstrates the DBDs of c-Jun and E47 do not form a complex with the DNA-bound PU.1 DBD. Together, these results suggest that either these interactions are too weak or transient to detect, or that other domains or proteins are crucial for interactions. Potentially weak or transient interactions were further investigated by use of a PU.1-GATA1 fusion protein, with the aim of structurally characterising any observed interactions by X-ray crystallography. Overall, this work demonstrates that PU.1 regulation of the M-CSFR promoter acts via mechanisms that are more complex than those described within the currently accepted model of binary TF interactions.

A PCR-RFLP method was used to identify cry2A toxin genes in a collection of 300 strains of Bacillus thuringiensis. From 81 genes identified, the vast majority appeared to be cry2Aa or cry2Ab, however three showed a different pattern and were subsequently cloned and sequenced. The gene cloned from strain HD395 was named cry2Ba2. Since the proteins encoded by the genes cloned from LS5115-3 and DS415 shared >95% sequence identity with existing toxins their genes were named cry2Aa17 and cry2Ab29 respectively by the toxin nomenclature committee. Despite this overall similarity these two toxins resembled natural hybrids, with Cry2Ab29 resembling Cry2Ab for the majority of the protein but then showing identity to Cry2Aa for the last 66 amino acids. For Cry2Aa17, Domains II and III most closely resembled Cry2Aa (99% identity) whilst Domain I was identical to that of Cry2Ab. The toxicity of the recombinant toxins was tested against Aedes aegypti and Spodoptera exigua, and it was found that the toxicity profile of Cry2Aa17 more closely matched the profile of Cry2Ab than that of Cry2Aa, thus implicating Domain I in specificity determination. This association of Domain I with toxicity was confirmed when hybrids were made between Cry2Aa and Cry2Ab.