Structural and related studies on Mycobacterium tuberculosis pantothenate kinase
MetadataShow full item record
Pantothenate kinase (PanK) is a ubiquitous and essential enzyme that catalyses the first step in the universal Coenzyme A (CoA) biosynthetic pathway. In this step pantothenate is converted to phosphopantothentate, which subsequently forms CoA in four enzymatic steps. Three types of PanKs have been identified in bacteria, with variations in distribution, mechanism of regulation, cofactor requirement and affinity for substrates. As part of a major programme on mycobacterial proteins in our laboratory, studies on type I PanK from Mycobacterium tuberculosis (MtPanK) have been carried out previously. This investigation involved, apart from biochemical studies, structure determination of twenty-one independent crystals of binary and ternary complexes of MtPanK involving CoA, the ATP analogue AMPPCP, the GTP analogue GMPPCP, ADP, GDP, pantothenate, phosphopantothenate, citrate, pantothenol and nonyl pantothenamide. Analysis of these structures brought out the robustness of the mycobacterial PanK when compared to its E. coli homologue. It was observed that while the protein structure remained relatively rigid in all the MtPanK structures, the ligands exhibited substantial movement in the pre-formed pocket during the course of catalysis. This observation was unlike that seen in EcPanK, where the protein molecule underwent conformational changes during enzyme action. Also, the feedback inhibitor, CoA, showed a higher binding affinity to MtPanK compared to that to EcPanK. The differences exhibited by these homologous proteins despite sharing 52% sequence identity were surprising and merited further study. To this end, mutants of MtPanK were prepared, their structures solved, and solution studies related to binding and activity of these mutants were carried out. Apart from this, supplemental molecular dynamics (MD) studies were carried out on MtPanK and EcPanK and the mutants of MtPanK. Structural studies were carried out using conventional tools and techniques of macromolecular crystallography. The microbatch under-oil method was used for crystallisation in all cases. Data were collected at a home-source on a MAR 345 image plate mounted on a Bruker MICROSTAR ULTRA II Cu Kα rotating-anode X-ray generator or using a CCD detector (MARMosaic 225) on the synchrotron X-ray beamline BM14 at the European Synchrotron Radiation Facility, Grenoble, France. Data were processed using MOSFLM and SCALA and the structures were solved by the molecular replacement method using PHASER from the CCP4 suite. Refinement was carried out using REFMAC and manual model building was performed employing COOT. Structures were validated using PROCHECK. Thermal shift assay was used to study binding of CoA to the mutants. A radioactive assay and an enzyme coupled assay were employed to measure the activity of the mutants. To begin with, the high affinity of CoA to MtPanK was sought to be disturbed by disrupting the binding site using mutations. Therefore, two conserved phenylalanine residues of the hydrophobic binding site were targeted and two point mutants and a double mutant were constructed. Solution studies on the three mutants confirmed the reduction in CoA binding affinity and also that of activity to some extent. Structure solution of the mutants showed that apart from local rearrangements, the mutations led to partial or complete transition of the structure to that seen in EcPanK. Concerted movement was observed in the dimerisation region and the nucleotide binding region. To further understand this transformation and as a complementary effort to the studies on the CoA binding region mutants, mutations were made in the substrate binding regions of MtPanK such that non-conservatively substituted residues were replaced with those found in EcPanK. Solution studies on these mutants showed that CoA binding affinity was minimally affected by the mutations, while activity was reduced to some extent. In all, six structures were solved, half of which were CoA-free and showed partial or complete transformation to an Ec-like conformational state. Concerted movement was seen in another loop along with that seen in the dimerisation interface and the nucleotide binding region. The structures brought to light the changes in conformations of certain residues and their interactions that makes the Ec-like state feasible in MtPanK. Put together, these studies showed how small perturbations like those caused by point mutations could bring about global transformations in the structure of MtPanK. They also suggest that MtPanK may be able to utilise the nucleotide binding pocket as seen in EcPanK in the transformed structures. This is an aspect that may be important in relation to drug designing. The results obtained from the mutational studies were supplemented by MD simulations on MtPanK, EcPanK and mutants of MtPanK. These studies helped delineate an invariant core common to MtPanK, EcPanK and the MtPanK mutants. They also showed that wild-type EcPanK is indeed more flexible than wild-type MtPanK. Furthermore, MD simulations showed the impact of sequence on the structure of the Mt enzyme. Minor sequence changes appeared to influence different structural elements, including those far away from the sites of mutation. Thus, it would seem that an ensemble of structures is accessible to the PanK molecule and the selection of an appropriate conformation is based on the requirement brought about by mutations or ligand binding. Apart from the studies on MtPanK, structural studies on argininosuccinate lyase from Mycobacterium tuberculosis were also carried out. The native structure along with that bound to the substrate and products helped propose a catalytic mechanism based on previously available information and present studies. This investigation is presented in an appendix.