The enzyme α-isopropylmalate synthase (IPMS) catalyses the reaction between acetyl coenzyme A (AcCoA) and α-ketoisovalerate (KIV) to produce free coenzyme A and α isopropylmalate (IPM). This reaction is a key control point in the biosynthesis of a leucine, a pathway absent in animals but present in plants, fungi and bacteria. As a result, IPMS is a antibiotic and herbicidal target that has been validated by knockout studies for M. tuberculosis, the causative agent of tuberculosis. Engineered IPMSs have also been used in the fermentative production of long chain alcohols for use as fuels. IPMS belongs to a family of related enzymes called α-ketoacid: AcCoA re-aldolases (KARAs), with each subfamily differing in the specific α-ketoacid that AcCoA is reacted with. The known KARA subfamilies are IPMS, citramalate synthases (CMSs), homocitrate synthases (HCSs), methylthioalkylmalate synthases (MAMSs) and re-citrate synthases (RCSs), respectively involved in the biosynthesis of isoleucine, lysine, glucosinolates and TCA cycle intermediates. This thesis describes work aimed at improving understanding of both specific subfamilies of KARA enzymes and also the genetic and functional relationships between the subfamilies. A particular emphasis is placed on relating primary structure to function, allowing the inference of function from a very small subset of residues. IPMSs are divided into two classes, the Mtu-like IPMSs and the much less studied Eco-like IPMSs. Chapter 2 details the expression and characterisation of the Eco like IPMS from N. meningitidis (NmeIPMS). Overall NmeIPMS showed similar properties to MtuIPMS, but unlike that enzyme NmeIPMS is inhibited by high divalent metal ion concentrations, does not require monovalent metal ions, and shows some activity with the α-ketoacid 3-methyl α ketovalerate. Several previous results showing inhibitory activity of Zn2+, Cd2+ and bromopyruvate were also found to be the results of interference with the assay system and all three were found to be much weaker inhibitors than previously determined.   Phylogenetic analysis of the different KARA subfamilies revealed certain specific positions that were believed to control substrate specificity. Chapter 3 details mutagenesis experiments on MtuIPMS that probe the function of these residues. Once the importance of the residues had been established, substitutions were made in which IPMS residues were replaced with their equivalents from HCSs and CMSs in order to change substrate specificity. The most successful result was the Y169L substitution based on HCS, which decreased the specificity constant with KIV by four orders of magnitude while improving other activities, successfully shifting the best activity to the unbranched α-ketoacid α-ketobutyrate. Chapter 4 of this thesis details the purification and functional testing of the RCS from C. saccharolyticus (CscRCS), the first thermophilic RCS characterised. CscRCS was found to have an extremely low Km for its substrate oxaloacetate (1.7 µM), believed to be an adaptation to the instability of oxaloacetate at the temperatures CscRCS operates at in vivo. The enzyme also showed competitive affinity by α-ketoglutarate, the end product of the pathway. Unlike other characterised RCSs, CscRCS showed no oxygen sensitivity. The phylogenetic analysis conducted for this thesis identified a subfamily of KARAs dubbed pseudo-IPMSs (PIPMSs) that showed no substantial homology to any studied subfamily. In Chapter 5 the PIPMS from T. thermophilus (TthPIPMS) was expressed and characterised. TthPIPMS showed many features of a CMS, being most active with the same substrate (pyruvate) and sensitive to the same inhibitor (isoleucine). Unlike the previously studied CMS subfamilies, TthPIPMS possesses a nanomolar IC50 for its inhibitor, and also shows substantial activity as an RCS. The results of these chapters are then drawn together in Chapter 6 to create a picture of the relationships between the KARA enzymes, in terms of their functional characteristics as well as the sequence and evolutionary relationships between them that have bought about their diverse functions.

α-Isopropylmalate synthase (α-IPMS) catalyses the first committed step in leucine biosynthesis in bacteria, including Neisseria meningitidis and Mycobacterium tuberculosis. It catalyses the condensation of α-ketoisovalerate (α-KIV) and acetyl coenzyme A (AcCoA) to form α-isopropylmalate (α-IPM). Like many key enzymes in biosynthesis, α-IPMS is inhibited by the end-product of the biosynthetic pathway, in this case leucine. α-IPMS is homodimeric, with monomers consisting of a (β/α)8-barrel catalytic domain, two subdomains and a C-terminal regulatory domain, responsible for binding leucine and providing feedback inhibition for leucine biosynthesis. The exact mechanism of feedback inhibition in this enzyme is unknown, despite the elucidation of crystal structures with and without leucine bound. This thesis explores the nature of allosteric regulation in α-IPMS, including the effects of the regulatory domain and the importance of structural asymmetry on catalytic activity. Chapter 2 details the characterisation of wild-type α-IPMS from N. meningitidis (NmeIPMS). This protein was successfully cloned, expressed and purified by metal-affinity and size-exclusion chromatography. NmeIPMS has similar characteristics to previously characterised α-IPMSs, being a dimer and demonstrating substrate binding affinities in the micromolar range. This enzyme has a turnover number of 13s⁻¹ and is sensitive to mixed, non-competitive inhibition by the amino acid leucine. Small angle X-ray scattering experiments reveal that the solution-phase structure of this protein is likely similar to existing crystal structures of other α-IPMSs. In Chapter 3, substitutions of residues potentially involved in the binding and transmission of the leucine regulatory mechanism are described. Most of these amino acid substituted variants reduce enzyme sensitivity to leucine, and one variant is almost entirely insensitive to this inhibitor. Another of these variants demonstrates an unexpected decrease in substrate affinity, despite the substituted residue being located far from the active site. The independence of α-IPMS domains is investigated in Chapter 4. The catalytic domains were isolated from NmeIPMS and the α-IPMS from M. tuberculosis (MtuIPMS), and found to be unable to catalyse the condensation of substrates, despite maintaining the wild-type structural fold. Complementation studies with Escherichia coli cells lacking the gene for α-IPMS show that the truncated variants are unable to rescue growth in these cells. Binding of α-KIV in the truncated NmeIPMS variant is much stronger than in the wild-type, and this may be the reason for lack of competent catalysis. A crystal structure was solved for the truncated variant of NmeIPMS and indicates that the regulatory domain is required for proper positioning of large regions of the protein. Two isolated regulatory domains from NmeIPMS were cloned, but with limited success in characterisation. Finally, Chapter 5 describes substitutions made in MtuIPMS to affect relative domain orientations within the protein. Dimer asymmetry is investigated by substituting residues at the domain interfaces. These substitutions did have some effect on catalysis and inhibition, but did not show any change in average solution-phase structure. These results are drawn together in the greater context of allostery in general in Chapter 6, along with ideas for future research in this field. This chapter reviews the insights gained into protein structure from this thesis, particularly the importance of residues at protein domain interfaces. The asymmetry in the α-IPMS structure is discussed, along with small-molecule binding regulatory domains.