Identification of gene targets against dormant phase Mycobacterium tuberculosis infections

Abstract

Background: Mycobacterium tuberculosis, the causative agent of tuberculosis (TB), infects approximately 2 billion people worldwide and is the leading cause of mortality due to infectious disease. Current TB therapy involves a regimen of four antibiotics taken over a six month period. Patient compliance, cost of drugs and increasing incidence of drug resistant M. tuberculosis strains have added urgency to the development of novel TB therapies. Eradication of TB is affected by the ability of the bacterium to survive up to decades in a dormant state primarily in hypoxic granulomas in the lung and to cause recurrent infections.

Methods: The availability of M. tuberculosis genome-wide DNA microarrays has lead to the publication of several gene expression studies under simulated dormancy conditions. However, no single model best replicates the conditions of human pathogenicity. In order to identify novel TB drug targets, we performed a meta-analysis of multiple published datasets from gene expression DNA microarray experiments that modeled infection leading to and including the dormant state, along with data from genome-wide insertional mutagenesis that examined gene essentiality.

Results: Based on the analysis of these data sets following normalization, several genome wide trends were identified and used to guide the selection of targets for therapeutic development. The trends included the significant up-regulation of genes controlled by devR, down-regulation of protein and ATP synthesis, and the adaptation of two-carbon metabolism to the hypoxic and nutrient limited environment of the granuloma. Promising targets for drug discovery were several regulatory elements (devR/devS, relA, mprAB), enzymes involved in redox balance and respiration, sulfur transport and fixation, pantothenate, isoprene, and NAD biosynthesis. The advantages and liabilities of each target are discussed in the context of enzymology, bacterial pathways, target tractability, and drug development.

Conclusion: Based on our bioinformatics analysis and additional discussion of in-depth biological rationale, several novel anti-TB targets have been proposed as potential opportunities to improve present therapeutic treatments for this disease.

Figure 1

Flowchart of the process used to generate the prioritized list of tractable therapeutic targets.

Figure 2

Overlap of the top 400 highest-scoring genes (~10% of the genome) from each of the three types of experimental models of dormancy. Murine refers to M. tuberculosis cells isolated from mouse macrophages, subcutaneous hollow fiber, and lung.

Figure 3

Ratio of the number of genes in the highest-scoring fraction (top 10%) to the number in the entire genome, for each classification described in Cole, et al. [17]. Ratios greater than 1 indicate that the genes are over-represented in the highest-scoring fraction relative to their representation in the whole genome.

Figure 4

Histogram of the distribution of upregulated and downregulated gene scores for the entire genome overlaid with the distribution from the devR regulon (A) and the relA regulon (B).

Figure 5

The two biosynthetic pathways for isopentenyl-pyrophosphate biosynthesis. The gene products (enzymes) from the M. tuberculosis genome presumed to catalyze each transformation are specified by the Rv numbers. Enzymes for which there is not a relevant orthologue in the M. tuberculosis genome are labeled as missing. [119] [76].