Exploring hot spot mutagenesis as a means of dimer disruption in glutamate racemase from Mycobacterium smegmatis

Abstract

The emergence of multidrug-resistant Mycobacterium tuberculosis strains has highlighted the need to develop new drugs for the treatment of tuberculosis. Glutamate racemase (MurI) has been identified as an essential enzyme in M. tuberculosis cell wall formation and has potential as a target for anti-tuberculosis drug development. Biochemical and structural information has recently been reported for MurI from M. tuberculosis (MurIMtb) and, a close homolog, M. smegmatis (MurIMsm). The structures of MurIMtb and MurIMsm were recently identified as homodimeric in solution, and both enzymes were reported as inactive in vitro. Mutagenesis, directed at disrupting the dimeric interaction, restored partial activity to MurIMsm and measurable amounts of monomeric species were detected in samples of the mutated protein. These results suggested that monomeric species of MurIMsm were responsible for the activity reported and it was, therefore, hypothesised that monomeric MurIMsm, engineered through dimer disruption, would show activity in vitro. The successful generation by mutagenesis could be applied to MurIMtb for use in the drug development process. This thesis describes the identification of specific dimer interface residues in MurIMsm and MurIMtb, known as hot spots, which account for most of the binding free energy of the respective dimers. A selection of computational prediction webservers were used to predict hot spots in the MurIMtb and MurIMsm protein structures. The goal of this in silico hot spot prediction was to identify residues that, when mutated, would disrupt dimeric interactions and produce potentially active monomeric protein. The comparative results from the prediction webservers identified eight hot spot residues in the MurIMsm structure, and the corresponding alanine substitution mutations for six of these were generated in MurIMsm. The MurIMsm mutant proteins were recombinantly expressed, purified, verified, and then oligomeric state and activity were determined. These mutant proteins were all identified as likely predominantly dimeric in solution. The mutant enzymes also showed varying amounts of activity ranging from zero, as observed with wild type MurIMsm, to activity comparable to that observed with previous MurIMsm dimer interface mutants.