Retracing the evolution of enzyme regulation.

α-Isopropylmalate synthase (IPMS) catalyses the first committed step in the leucine biosynthesis pathway in microorganisms and some plants. It catalyses the condensation of ketoisovalerate (KIV) and acetyl-coenzyme A (AcCoA) to form isopropylmalate and coenzyme A (CoA). IPMS is allosterically inhibited by the product of the pathway, L-leucine. Structurally, IPMS is a homodimer, and each chain consists of a N-terminal (α/β)8 barrel where the active site is located, a catalytic accessory unit formed of subdomain I and subdomain II, and a C-terminal regulatory domain that binds L-leucine. Truncation of IPMS that removes subdomain II or part of subdomain II abolishes catalysis.¹,² Of particular interest in this thesis is IPMS from Neisseria meningitidis (NmeIPMS). Although this enzyme has been extensively studied, there is no full-length crystal structure available, although there are several of a related enzyme, IPMS from Mycobacterium tuberculosis (MtuIPMS), with KIV, and with L-leucine, bound. There is no substantial conformational change observed when the substrate-bound crystal structure is compared to the structure with L-leucine bound. Untangling the dynamic nature of allostery, and how it has evolved, in these proteins is a particular focus of this thesis. There are also structurally similar proteins that catalyse similar reactions that are also of interest. Citramalate synthase (CMS) is structurally similar to IPMS but catalyses the reaction of pyruvate and AcCoA to form citramalate and CoA in an isoleucine biosynthesis pathway and is inhibited by L-isoleucine. Homocitrate synthase (HCS) functions in a lysine biosynthesis pathway in some organisms and utilises ketoglutarate and AcCoA to form homocitrate and CoA. HCS contains a homologous catalytic domain and catalytic accessory unit as IPMS but lacks a regulatory domain and is competitively inhibited by lysine. The similarities, differences, and modularity of these proteins is explored using computational methods and also by the construction of truncated and fusion proteins. Chapter 2 utilises a computational method, statistical coupling analysis, to identify a potential network in NmeIPMS-like IPMS proteins. Subsequent alanine mutations in NmeIPMS demonstrated that mutating charged residues in this proposed network can abolish or attenuate the allosteric signal, suggesting that the network identified may represent a way the allosteric signal is transferred from the allosteric site to the active site. Isothermal titration calorimetry is also used to explore the thermodynamics of L-leucine binding to the wild-type NmeIPMS and to the L-leucine insensitive alanine mutants. Chapter 3 broadens the scope of statistical coupling analysis (SCA) and also uses another computational method, mutual information (MI), to investigate how structurally similar subdomains facilitate catalysis in the presence and absence of a regulatory domain. A population of proteins that contain a regulatory domain, and a population that do not, were assessed using SCA and MI to determine whether there were differences, particularly in the subdomains, that may provide information about maintenance of the balance of flexibility and stability that is crucial to catalysis in these proteins. Chapter 4 uses an active, truncated, form of NmeIPMS to compare and contrast with the wild-type protein. The kinetics of both the truncated NmeIPMS and the wild-type NmeIPMS are investigated under crowded conditions to explore the impact that viscosity has on these dynamic proteins. Alanine mutations are also made in subdomains I and II to investigate the role of particular residues in catalysis and allostery, and these allow comparison with previous work done on MtuIPMS that highlights the difference between two structurally similar groups of IPMS proteins. Chapter 5 describes the cloning, expression, and partial purification of an HCS, SsoHCS, from Sulfolobus solfataricus that appears to have a different type of regulatory domain to the canonical IPMS/CMS regulatory domain. The partial characterisation of this protein suggests that an allosterically regulated HCS has been identified. Chapter 5 also describes the construction of several fusion proteins, where parts of IPMS, HCS, and CMS, are fused to together to explore the modularity of these proteins. Catalysis was preserved in some of the fusions although allostery was not preserved in any so far investigated. The final chapter includes a broad summary of the work in this thesis as well as ideas for future research. This chapter also contains a discussion about the considerable differences between some IPMS enzymes that, although they catalyse the same reaction, are considerably different taxonomically. Additionally, the important role of networks of residues that facilitate catalysis and allostery is analysed.