The chemistry of the tetrameric low-valent aluminium compoud (Cp*Al)4 (Cp* = 1,2,3,4,5- pentamethylcyclopentadienyl) is relatively undeveloped compared to its monomeric cousin dippNacNacAl (dippNacNac = 2,6-diisopropylphenyl-β-diketiminate). Given that the former can be formed by the reductive elimination of Cp*H from Cp*2AlH, a process common to transition metals yet rare with light main-group elements, using the Cp* ligand could unlock an abundance of unexpected reactivity for aluminium. An overview of the literature regarding the synthesis and reactivity of low oxidation state aluminium compounds is provided in chapter 1, as well as an introduction to relevant magnesium chemistry for this work. Chapter 2 studies the mechanism of C-H reductive elimination from Cp*2AlH to form (Cp*Al)4, and the properties which allow reductive elimination to take place are revealed. A transition state is identified where the Cp* group has a higher hapticity than in the starting material, a process which is thought to enable the reductive elimination. Using this insight, aluminium hydride and halide complexes featuring 9-methylfluorenyl ligands are synthesised and reduction of the aluminium centre is investigated. The reactivity of (Cp*Al)4 is considered in chapter 3 of this thesis. The formal cycloaddition reaction between (Cp*Al)4 and diphenylacetylene produces a Lewis acidic 1,4- dialuminacylohexadiene derivative. The inner Al2C4 ring of this complex is stable, with onward reactions happening at the complex's periphery. Insertion reactions in the Al-CCp* bonds are observed with unsaturated C-N species. With 2,6-dimethylphenylisonitrile the Al2C4 complex forms a zwitterionic aluminate, featuring a stable carbocation derived from the Cp* group. An amidinate complex with an unusual Cp* backbone is formed from the insertion of carbodiimides into the Al-CCp* bond of the 1,4-dialuminacyclohexadiene. Extending this, the insertion of carbon dioxide into the same bond is explored. The use of amidine ligands is common in main-group chemistry, however literature relating to the related phosphaamidinate ligands ([RPC(R)NR]-) is only reported sporadically. They have not been applied in a general manner to main-group chemistry thus far. Chapter 4 describes the synthesis of five new phosphaamidinate pro-ligands where the steric bulk of both the phosphorus and nitrogen components is increased systematically. To evaluate these new ligands, their coordination chemistry with magnesium was investigated. Three examples of heteroleptic LMgnBu (L = phosphaamidinate) complexes are synthesised, which all show high activity for the ring-opening polymerisation of racemic lactide. The resulting polylactide chains show good molecular weights and polydispersity indices. The synthesis of homoleptic L2Mg complexes is also described. Chapter 5 applies these new phosphaamidinate ligands to aluminium chemistry. An aluminium hydride species is isolated, which is shown to form via a probable lithium aluminate intermediate. The lifetime of this intermediate is found to be heavily dependent on the reaction solvent.

Transcriptional deregulation has emerged as one of the leading causes in various human diseases. More than fifty percent of cancers arise due to frequent mutations in genes regulating transcription. Higher-order assembly via protein-protein interactions is one common mechanism of transcriptional regulation. Despite their critical role in regulating gene transcription and therapeutic relevance, detailed mechanistic understanding of these assemblies remains scarce. The primary focus of this thesis is to uncover important structural principles underlying the assembly and stability of multi-domain protein assemblies by characterizing components of the IFNβ enhanceosome and the CtBP-mediated repression complex. Using a combination of biochemical and structural analyses, I showed that the transcriptional activator C-terminal binding protein 2 (CtBP2) is a tetramer by solving the 3.6Å cryoEM structure of CtBP2. I highlighted the types of interactions that stabilize the homo-tetramer and showed the relevance of the tetramer in CtBP2 transcriptional activity. Second, I used an integrative approach to investigate the structural features leading to activation of interferon regulator factor 3 (IRF3) and its interaction with DNA. Although this work mostly focused on components of the CtBP2-mediated complex and IFNβ enhanceosome, the principles described here can be applied to other complexes. Therefore, these studies provide an overall understanding on how other macromolecular complexes regulate gene transcription. Furthermore, our structural-based analyses will provide a basis for designing drugs that can regulate gene transcription in cancer and immunological disorders.

<p> We developed some universal platforms to overexpress the minimal functional entities of ion channels. The modular property of ion channels have been demonstrated from many aspects, such as crystal structures, chimeric channel experiments and discovery of similar modules in distantly related protein families. Thus it should be feasible to express each module independent of other channel modules. The pore-forming module of ion channels has multiple important properties as selectivity, conductivity and drug-binding. If it can be overexpressed, it will provide valuable information about channel selectivity to different ions and structural bases for drug binding as well as important application in drug screening and rational drug design. </p> <p> To test this, we first used the model channel KcsA to identify the minimal requirements for a pore-forming domain to functionally exist independently. Chapter 2 of this thesis explains in detail how the wild type C-terminal cytoplasmic domain of KcsA functions. We found that this domain has dual function as pH-sensor and tetramerization domain, and it is essential for the expression of the pore-forming domain of KcsA. Once we knew the physiological role of the cytoplasmic domain, the scenario was set to answer the question of how to make it better for the application of structural and functional studies. </p> <p> In chapter 3 and chapter 4, we replaced the wild type C-terminal domain with non-native tetramerization domains. We identified the direct correlation between protein expression level and overall thermostability of pore-forming domains. The C-terminal tetramerization domains stabilize channels in a cooperative way and play a critical way in in vivo channel assembly. The selection of the linker between pore-forming domain and tetramerization domain, the splicing motif, and the handedness of C-terminal tetrameric coiled coils all affect channel expression level and stability. </p> <p> We applied our finding in KcsA to a wide range of ion channels in chapter 5, including voltage-gated potassium channels, Ca2+-gated potassium channels, inwardrectifying potassium channels, cyclic nucleotide-gated potassium channels and voltagegated sodium channels. We managed to express similar minimal structural modules from these more structurally complicated channels with the assistance of different cytoplasmic tetramerization domains. Several minimal channels expressed well and showed similar biophysical and functional property as the wild type channels. </p> <p> These studies demonstrate that the pore-forming modules of ion channels can be expressed independently while retaining the proper structure and drug-binding properties as their wild type predecessors when using our universal expression platform. The potential application in structural studies and drug-screening is promising. </p> / Thesis / Doctor of Philosophy (PhD)