Targeting the chromosome partitioning protein ParA in tuberculosis drug discovery

OBJECTIVE

To identify inhibitors of the essential chromosome partitioning protein ParA that are active against Mycobacterium tuberculosis.

METHODS

Antisense expression of the parA orthologue MSMEG_6939 was induced on the Mycobacterium smegmatis background. Screening of synthetic chemical libraries was performed to identify compounds with higher anti-mycobacterial activity in the presence of parA antisense. Differentially active compounds were validated for specific inhibition of purified ParA protein from M. tuberculosis (Rv3918c). ParA inhibitors were then characterized for their activity towards M. tuberculosis in vitro.

RESULTS

Under a number of culture conditions, parA antisense expression in M. smegmatis resulted in reduced growth. This effect on growth provided a basis for the detection of compounds that increased susceptibility to expression of parA antisense. Two compounds identified from library screening, phenoxybenzamine and octoclothepin, also inhibited the in vitro ATPase activity of ParA from M. tuberculosis. Structural in silico analyses predict that phenoxybenzamine and octoclothepin undergo interactions compatible with the active site of ParA. Octoclothepin exhibited significant bacteriostatic activity towards M. tuberculosis.

CONCLUSIONS

Our data support the use of whole-cell differential antisense screens for the discovery of inhibitors of specific anti-tubercular drug targets. Using this approach, we have identified an inhibitor of purified ParA and whole cells of M. tuberculosis.

Global analysis of proteins synthesized by Mycobacterium smegmatis provides direct evidence for physiological heterogeneity in stationary-phase cultures

We have characterized the induction kinetics of approximately 1,700 proteins during entry into and survival in carbon-starved stationary phase by Mycobacterium smegmatis. Strikingly, among the patterns of expression observed were a group of proteins that were expressed in exponential-phase cultures and severely repressed in 48-h stationary-phase cultures (Spr or stationary-phase-repressed proteins) but were synthesized again at high levels in > or =128-day stationary-phase cultures (Spr(128) proteins). A number of Spr(128) proteins were identified, and they included the heat shock protein DnaK, the tricarboxylic acid cycle enzyme succinyl coenzyme A synthase, a FixA-like flavoprotein, a single-stranded DNA binding protein, and elongation factor Tu (EF-Tu). The identification of EF-Tu as an Spr(128) protein is significant, as ribosomal components are known to be expressed in a growth rate-dependent way. We interpreted these data in terms of a model whereby stationary-phase mycobacteria comprise populations of cells that differ in both their growth status and gene expression patterns. To investigate this further, we constructed gene fusions between the rpsL gene promoter (which heads the Mycobacterium smegmatis operon encoding the tuf gene encoding EF-Tu) or the rrnA promoter gene and an unstable variant of green fluorescent protein. While the majority of cells in old stationary-phase cultures had low levels of fluorescence and so rpsL expression, a small but consistently observed population of approximately 1 in 1,000 cells was highly fluorescent. This indicates that a small fraction of the cells was expressing rpsL at high levels, and we argue that this represents the growing subpopulation of cells in stationary-phase cultures.

Tetracycline-inducible gene regulation in mycobacteria

A system for the tetracycline-inducible regulation of gene expression in mycobacteria has been developed. We have sub-cloned the tetRO region from the Corynebacterium glutamicum TetZ locus into a mycobacterial shuttle plasmid, making expression of genes cloned downstream of tetRO responsive to tetracycline. Using the luxAB-encoded luciferase from Vibrio harveyi as a reporter (pMind-Lx), we observed a 40-fold increase in light output from Mycobacterium smegmatis cultures 2 h after adding 20 ng ml⁻¹ of tetracycline. Similarly, exposure to the drug resulted in up to 20-fold increase in relative light units from M.bovis BCG carrying the reporter construct, and a 10-fold increase for M.tuberculosis. Tetracycline induction was demonstrated in log and stationary phase cultures. To evaluate whether this system is amenable to use in vivo, J774 macrophages were infected with M.bovis BCG[pMind-Lx], treated with amikacin to kill extracellular bacteria, and then incubated with tetracycline. A 10-fold increase in light output was measured after 24 h, indicating that intracellular bacteria are accessible and responsive to exogenously added tetracycline. To test the use of the tetracycline-inducible system for conditional gene silencing, mycobacteria were transformed with a pMind construct with tetRO driving expression of antisense RNA for the ftsZ gene. Bacterial cells containing the antisense construct formed filaments after 24 h exposure to tetracycline. These results demonstrate the potential of this tetracycline-regulated system for the manipulation of mycobacterial gene expression inside and outside cells.

Development and application of unstable GFP variants to kinetic studies of mycobacterial gene expression

Unstable variants of green fluorescent protein (GFP) tagged with C-terminal extensions, which are targets for a tail specific protease, have been described in Escherichia coli and Pseudomonas putida [Appl. Envir. Microbiol. 64 (1998) 2240]. We investigated whether similar modifications to flow cytometer optimised GFP (GFPmut2) could be used to generate unstable variants of GFP for gene expression studies in mycobacteria. We constructed GFP variants in a mycobacterial shuttle vector under the control of the regulatory region of the inducible Mycobacterium smegmatis acetamidase gene. GFP expression was induced by the addition of acetamide and the stability of the GFP variants in M. smegmatis, following the removal of the inducer to switch off their expression, was determined using spectrofluorometry and flow cytometry. We demonstrate that, compared to the GFPmut2 (half-lives>7 days), the modified GFP variants exhibit much lower half-lives (between 70 and 165 min) in M. smegmatis. To investigate their utility in the measurement of mycobacterial gene expression, we cloned the promoter region of a putative amino acid efflux pump gene, lysE (Rv1986), from Mycobacterium tuberculosis together with the divergently transcribed, putative lysR-type regulator gene (Rv1985c) upstream of one of the unstable GFP variants. We found that the expression kinetics of the lysRE-gfp fusion were identical throughout the M. smegmatis growth curve to those measured using a conventional lysRE-xylE reporter fusion, peaking upon entry into stationary phase. In addition, it was established that the tagged GFP variants were also unstable in Mycobacterium bovis BCG. Thus, we have demonstrated that unstable GFP variants are suitable reporter genes for monitoring transient gene expression in fast- and slow-growing mycobacteria.