Pushing The Boundaries Of Target-based Tuberculosis Drug Discovery With Inhibitors Of The Essential Cell Division Gene, parA
Research For Life
School of Biological Sciences
New drugs are required to treat multi-drug resistant and latent forms of human tuberculosis, and to shorten the therapeutic time to below six months. The discovery of drugs is dependent on assays that can reliably identify anti-mycobacterial compounds in high throughput. In earlier work in our laboratory, we identified a series of potential inhibitors of the essential cell division gene parA using a targeted whole cell antisense screen. In this project, we set out to validate these compounds as specific inhibitors of ParA and furthermore, ascertain their potential against clinical strains of Mycobacterium tuberculosis. Specifically, the primary objectives of this project were:
A. Determine the dose response of Mycobacterium smegmatis to the ParA inhibitors.
The 90 % minimum inhibitory concentration (MIC90) values of the potential parA inhibitors, octoclothepin and phenoxybenzamine, towards M. smegmatis were determined with respect to controls. MIC90 values for octoclothepin ranged from 61.2 µM in nutrient-rich media to 38.2 µM in nitrogen-limited growth media. The MIC90 values of phenoxybenzamine ranged from > 1000 µM in nutrient-rich media to 132.2 µM in nitrogen-limited growth media. While the activity of more potent compound, octoclothepin, may increase only slightly under nitrogen limitation, the change in activity for phenoxybenzamine under nitrogen limitation is more significant.
These data have been published in: Nisa S, Blokpoel MCJ, Robertson BD, Tyndall JDA, Lun S, Bishai WR, O’Toole R* (2010). Targeting the chromosome partitioning protein ParA in tuberculosis drug discovery. Journal of Antimicrobial Chemotherapy 65(11): 2347-2358.
B. Predict the interactions of specific inhibitors with ParA in silico. A three-dimensional structure of the ParA protein of M. tuberculosis was constructed in silico based on the crystal structures of MinD and Soj. Low-energy ligand structures were generated either using OMEGA2 and protonated using filter (Openeye software) or were generated in SYBYL-X using the sketcher module. Flexible ligand docking was carried out with Gold 4.134 using the ChemScore scoring function. Both compounds, octoclothepin and phenoxybenzamine, were predicted to fit within the ATPase active site of ParA and the active site amino acids underlying favourable interactions with the inhibitors were identified.
C. Examine biochemically the effect of inhibitors on purified ParA protein.
The ParA protein of M. tuberculosis was expressed using the pET-28a(+) vector and E. coli strain BL21 (DE3) in conjunction with a HisBind purification kit (Novagen) for purification of the ParA protein. The effect of octoclothepin and phenoxybenzamine on the ATPase activity of M. tuberculosis ParA was determined in the presence and absence of cis-acting site pars through the measurement of pyrophosphate (PPi) release. Addition of octoclothepin resulted in an approximate 20 % decrease in ParA ATPase activity at a concentration of 25 µM. Phenoxybenzamine decreased the activity of ParA ATPase by approximately 30 % at a concentration of 100 µM. Clinical tuberculosis drugs streptomycin had no significant effect on the ATPase activity of ParA. The effect of octoclothepin and phenoxybenzamine in biochemical assays with ParA support the in silico and earlier antisense data that these compounds inhibit ParA in mycobacteria.
D. Determine the effect of ParA inhibitors on growth and virulence of Mycobacterium tuberculosis.
Dr O’Toole and his collaborators at Johns Hopkins University tested the ParA inhibitors in virulent strains of M. tuberculosis at the Center for Tuberculosis Research, Johns Hopkins University School of Medicine, Baltimore, USA. Octoclothepin was active against virulent strains with an MIC90 ranging from to 17.4 to 34.7 µM. Phenoxybenzamine, in contrast, showed low activity towards virulent M. tuberculosis. The findings with octoclothepin validate the use of the antisense differential sensitivity approach with regard to identifying inhibitors of essential targets in M. tuberculosis.
We are now also using differential antisense susceptibility testing to confirm the interaction of clinical tuberculosis drugs with secondary cellular targets. For example, we have found that an NADH pyrophosphatase enzyme, and a ferrodoxin uptake regulator, desensitise mycobacterial cells to the mechanism of action of the first-line drug, isoniazid. Through the use of chemical derivatisation, it may be possible to identify isoniazid analogues which overcome drug detoxification mechanisms in mycobacteria. This latter work will be published in due course.