Mycobacterium Tuberculosis Evolutionary Pathogenesis and its Putative Impact on Drug Development

Fabien Le Chevalier; Alessandro Cascioferro; Laleh Majlessi; Jean Louis Herrmann; Roland Brosch


Future Microbiol. 2014;9(8):969-985. 

In This Article

Potential Roles of Essential Genes as Drug Targets

The identification of a set of essential genes for M. tuberculosis has greatly enhanced the ability to define new, potentially vulnerable drug targets of the pathogen. In combination with the advances in crystallography and the increasing availability of structural information on mycobacterial proteins, in silico drug screening should be a promising way to find new active molecules against M. tuberculosis. However, this approach must also be taken with caution, as binding of a given compound to an essential target protein under experimental condition often does not translate to actual growth inhibition of the bacterium in subsequent MIC screens. PknB, a serine–threonine protein kinase of M. tuberculosis, represents a typical example of such a situation. This protein, which is part of the 11 eukaryotic-like serine–threonine protein kinases of M. tuberculosis,[11] is listed as an essential protein in the M. tuberculosis TraSH and NGS datasets.[29,31] The essentiality of this protein has also been suggested by independent genetic studies, in which the knockout of the original pknB gene only became possible once a second intact pknB copy was integrated into the M. tuberculosis genome.[32] Similarly, the essentiality of PknB was also demonstrated by gene knockdown experiments, which showed a dramatic impairment of bacterial viability in liquid culture upon depletion of the kinase.[33] Nevertheless, potential inhibitors that showed affinity to PknB were active only in the higher micromolar MIC range,[34,35] which makes their further development into potential anti-TB drugs rather difficult and/or unlikely. For proteins involved in the regulation of essential cell processes, very small amounts might already be sufficient, so that it is very difficult for an inhibitor to fully block the function of the protein.

Apart from potential quantitative particularities, many other factors may impact on the final inhibition activity of a compound on the bacterium. For M. tuberculosis, the complex, lipid-rich cell wall composition, which in addition is protected by a polysaccharide-based capsule,[15,18,19] certainly complicates permeability. In addition, the potential effect of efflux pumps and protein binding in the assay media may also play a role in the evaluation of the final antimicrobial activity of a compound. Thus, according to most recent trends in anti-TB drug development, whole-cell screening of compound libraries has re-emerged as a frequently used technique to find new active compounds against M. tuberculosis for which the drug target can then be identified in follow-up studies. This approach is of great importance for enlarging the arsenal of presently used first-line (isoniazid, rifampicin, ethambutol and pyrazinamide) and second-line (streptomycin, fluoroquinolones, para-aminosalicylic acid and the injectable agents amikacin, kanamycin or capreomycin) anti-TB drug regimens (Table 1 & Table 2)[36] that may all be subject to resistance mechanisms in extensively and/or totally resistant M. tuberculosis strains.[37]

Screening and target identification studies allowed the diarylquinoline drug TMC207 to be discovered, for example, for which resistant mutants showed single-nucleotide polymorphisms in the gene atpE encoding the C chain of the ATP synthase of M. tuberculosis.[39] The activity of TMC207 on multidrug-resistant M. tuberculosis was confirmed in clinical trials,[40] and in December 2012, the drug (also known as bedaquiline or Sirturo™ [Janssen Therapeutics, NJ, USA]) was approved by the US FDA as a new treatment for multidrug-resistant TB that can be used as an alternative when other drugs fail. Other recently discovered active compounds against M. tuberculosis with potential for further development are the benzothiazinones, which are extremely efficient at blocking the growth of M. tuberculosis under in vitro conditions and are also active in vivo in mouse models.[41] Benzothiazinones target DprE1, which is involved in the synthesis of arabinans, essential components of the mycobacterial cell wall. DprE1 seems to be a highly vulnerable target, as recently shown for a range of compounds with different chemical scaffolds.[38,42] For example, DprE1 was found as the target of dinitrobenzamide derivatives (DNref-1), a class of compounds that were identified by high-content screening to interfere with the replication of M. tuberculosis within macrophages.[43] The employed screening technique, which is based on automated, simultaneous imaging of fluorescent bacteria and host macrophages exposed to different representatives of chemical compound libraries, is a particularly attractive and powerful method for identifying molecules that show activity against intracellular M. tuberculosis but little or no toxicity towards host cells.[43] The same high-content screening approach was also the starting point for the discovery of Q203, a potent candidate drug that targets qcrB, encoding for an ubiquinol–cytochrome C reductase of the respiration chain of M. tuberculosis.[44]

Another vulnerable drug target that is hit by different classes of small-molecule inhibitors is MmpL3, an essential mycobacterial membrane protein that harbors 11 transmembrane domains and is crucial for the transport of cell wall constituents.[45] Probably the most advanced compound in terms of clinical development that was found to target MmpL3 is the diamine SQ109, a drug candidate that is in Phase II of clinical testing.[46] Another molecule that is in clinical studies is PA-824, a nitroimidazole compound,[47] which targets a deazaflavin-dependent nitroreductase.[48]

Most recently, thiophene compounds were identified that kill M. tuberculosis by the previously uncharacterized mechanism of inhibition of the polyketide synthase Pks13,[49] which is an essential enzyme that catalyzes the last condensation step of mycolic acid biosynthesis.[50] In addition, a series of fluoroquinalones were studied in preclinical and clinical settings, with a particular focus on gatifloxacin and moxifloxacin, in order to evaluate whether they can improve the activity of the standard drug regimen when substituted for ethambutol and thereby might help to shorten the duration of treatment for fully drug-susceptible TB.[51] As another example of the clinical evaluation of new drugs, the addition of linezolid into the regimen against extensively drug-resistant forms led to strongly improved prognosis of the treatment outcomes and culture conversion in patients with severe, chronic, extensively drug-resistant TB,[52] which gives hope that other oxazolidinone-based drugs currently under development might also be effective.[53]

Recent research has also elucidated the targets and putative modes of action of some anti-TB drugs that have been known for a long time. One striking example is the identification of the enoyl reductase InhA as the drug target of the natural compound pyridomycin,[54] which bridges the NADH- and substrate-binding pockets of the enzyme and thus inhibits InhA in a different way than the first-line drug isoniazid, which is a prodrug and needs activation by KatG in order to form the active isonicotinic acyl–NADH complex. As a result, pyridomycin is also active on the most frequently encountered isoniazid-resistant M. tuberculosis strains, which opens new perspectives for drug development.[55] Another old drug that targets InhA is ethionamide, a thiocarbamide-containing compound that is activated by the mycobacterial monooxygenase EthA, whose production is controlled by the transcriptional repressor EthR. A recently applied strategy that is based on inhibiting EthR shows promise for improving the therapeutic index of thiocarbamide derivatives and for overcoming resistance.[56] Finally, for two other 'old' drugs, pyrazinamide and para-aminosalicylic acid, new mechanisms of action were recently identified. For pyrazinamide, activity directed against persistent forms of M. tuberculosis was discerned, which involves the inhibition of the process of trans-translation by ribosomal protein S1, which is essential for freeing scarce ribosomes in nonreplicating organisms.[57] For para-aminosalicylic acid, it was found that it acted as a prodrug, which poisons folate-dependent pathways not only by serving as a replacement substrate for dihydropteroate synthase, but also by the generation of toxic byproducts created by the enzymes of these pathways.[58]