Abstract:
Mycobacterium tuberculosis (Mtb) is the causative agent of tuberculosis (TB). In 2012, TB was responsible for 1.7 million deaths worldwide, making Mtb one of the most devastating human pathogens. The bacillus has an extraordinary capacity to adapt to low oxygen conditions by slowing down its growth in a phenomenon commonly, referred to as non-replicating persistence, giving rise to latent TB infections. The poor efficiency of current therapies against latent TB, together with the emergence and spread of multidrug- and extensively drug-resistant strains of the bacterium, make the discovery of new drug targets a high priority. The metabolism of amino acids, or the pathways to their biosynthesis, can provide promising leads towards development of new vaccines and/or anti-TB agents. Despite implications of various amino acids in Mtb survival and pathogenesis, none of the current anti-TB drugs or agents currently in clinical trials targets amino acid metabolism or biosynthesis. In this project, we have focused on amino acids that are either not synthesised by humans, such as branched chain amino acids (BCAAs) or whose presence provides an advantage during Mtb pathogenesis (i.e. proline). In the first part of this PhD project, the Mtb dihydroxyacid dehydratase enzyme was targeted for structural studies. This enzyme catalyses the third step in the BCAA biosynthesis pathway and is a putative [Fe-S] cluster-binding protein. The wild type protein, as well as five surface mutants, were expressed and purified. All gave brown protein solutions, implying that an [Fe-S] cluster was present. This protein was submitted to crystallisation trials. Colourless crystals were obtained for the wild type and two of the surface mutant constructs, all of which diffracted to ~3.5 Å resolution. Poor and mosaic diffraction has so far hindered determination of the threedimensional structure of the protein. This could, at least in part, be attributed to protein heterogeneity resulting from degradation and/or loss of the [Fe-S] cluster. The second part of this project aimed to investigate the enzymes involved in conversion of proline to glutamate (PruA and PruB), along with the transcriptional regulator of these enzymes (PruC). All three enzymes were successfully expressed and purified in either E. coli or M. smegmatis expression systems. Despite numerous efforts, the PruB enzyme was consistently found to be aggregated and thus remained unsuitable for structural studies. On the other hand, PruA was obtained as a soluble homodimer and was successfully crystallised. Two crystal forms of PruA were obtained, in one of which the unit cell dimension doubled along the c axis (194 against 96 Å). Data collected from crystals with the long unit cell dimension could be processed in both P622 and P6322 hexagonal space groups, implicating some lattice dislocation defects. The structure of SeMet-substituted PruA was solved using the multi-wavelength anomalous dispersion technique and refined at 2.5 Å resolution, with associated crystallographic R and Rfree factors of 14.4 and 20.7%, respectively. The native PruA structure was then solved by molecular replacement at 2.04 Å resolution (R=22.9% and Rfree=24.1%). The PruA-NAD+ complex structure was solved at 2.25 Å resolution (long unit cell) and 1.6 Å resolution (short unit cell) with R/Rfree factors of 27.7/30.7% and 17.1/20.2%, respectively. In an attempt to overcome the observed lattice dislocation defects, Mtb-PruA was co-crystallised with cobalamin through screening with the Silver Bullets Bio screen. The complex structures of PruAcobalamin and PruA-cobalamin-NAD+ were solved at 2.27 Å resolution (R=14.8% and Rfree=20.4%) and 2.38 Å resolution (R=14.9% and Rfree=20.1%), respectively. The binding of cobalamin resulted in new crystal packing in space group P3121, a form that lacked the aforementioned crystallographic complications. The PruA active site shows a high degree of conservation when compared with homologous structures, but it contains some specific features that could be exploited for drug design attempts. The PruA and PruB enzymes were also used to investigate the conversion of proline into glutamate. A combination of NMR and UV-visible spectroscopy allowed simultaneous monitoring of proline consumption and formation of glutamate. PruC was then expressed and purified in E. coli as a fusion protein with maltose binding protein (MBP), although the protein was aggregated in its absence. This research provides some new insights into amino acid metabolism in Mtb. The results of the Mtb-DHAD study are encouraging and suggest the possibility of further improvement that could enable determination of its three-dimensional structure, which would constitute an important step forward towards structure-based drug development. Our results also shed light on the proline utilisation pathway in Mtb, with functional and structural characterisation of the two enzymes involved in the pathway. The results of this research could be potentially used as foundation for the development of Mtb-PruA inhibitors and potential novel anti-TB agents.