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

Selected Virulence Factors of M. Tuberculosis & Their Potential Role for Alternative Intervention Strategies

As a result of the significant advances in mycobacterial molecular genetics and genomics in recent years, a number of M. tuberculosis genes that intervene in host–pathogen interactions have been identified. The first such mycobacterial virulence genes were identified by the use of pioneering gene knockout techniques[76,77] that were later combined with signature-tagged mutagenesis,[78,79] resulting in a first list of candidate genes involved in the pathogenicity of M. tuberculosis. Among the genes identified by this approach were several with insertions in a 70-kb chromosomal segment that encodes proteins involved in the synthesis and transport of phthiocerol dimycocerosate, which corresponds to specific, extractable lipids in the outer layers of the mycobacterial cell envelope that play an important role in virulence. The number of potential virulence genes was later extended by the use of the TraSH technique adapted to mouse infection, which predicts that, in addition to the 15% of genes that are essential for in vitro growth, another 5% are involved in in vivo growth.[30] Among the approximately 200 listed genes are also several that encode proteins that belong to macromolecular components or secreted proteins of the aforementioned major secretion systems of M. tuberculosis, which might be particularly interesting as potential drug targets, as their encoded proteins are supposed to be localized in the cell envelope and thus more easily accessible by putative inhibitors. As a first example, the gene secA2, which codes for a main constituent of the alternative SecA2-operated pathway, is part of this list. This pathway was shown to export SodA and KatG, which are both involved in the detoxification of reactive oxygen intermediates produced by the host cell as part of the oxidative attack within the phagosome.[69,80] Small-molecule inhibitors that would inhibit these oxidative stress defense mechanisms of M. tuberculosis might strongly impact on the fitness and transmissibility of the pathogen, as suggested by virulence studies on KatG-deletion mutants in the context of isoniazid resistance.[81,82]

The list of mutants with in vivo growth defects also points to multiple genes implicated in the ESX-1 type VII secretion pathway.[30] The ESX-1 secretion system was identified as a primary factor determining the host–pathogen interactions of tubercle bacilli by several independent and complementary approaches (reviewed in[61]). Indeed, deletion of part of the ESX-1-encoding chromosomal region, named region of difference 1 (RD1), from Mycobacterium bovis BCG, which still represents the only anti-TB vaccine currently used on a global scale, is one of the main reasons for the loss of virulence of this attenuated live vaccine. Similarly, the ESX-1 system is also truncated in Mycobacterium microti strains, of which several were used as live-attenuated vaccines in the 1960s.[61] Moreover, the presence of the 6-kDa early secreted antigenic target ESAT-6 and its protein partner CFP-10, which are both secreted by ESX-1 in M. tuberculosis, but are absent from M. bovis BCG and M. microti-based vaccines, represents the basis for the differential potential of the next-generation IFN-γ release assays that have substantially refined TB diagnosis.[83]

ESX-1 might thus serve as a potential target for the development of molecules that could interfere with the process of ESX-1-mediated host–pathogen interactions. Despite intense research within the last decade since its discovery, the exact function of the ESX-1 system is only partially known. However, recent pathogenicity-related research has shown that the intracellular behavior of wild-type M. tuberculosis is profoundly different from ESX-1-deficient variants of M. tuberculosis (ΔRD1) or other ESX-1-deleted tubercle bacilli, such as BCG. The main differences observed during the infection of macrophages are linked to the finding that ESX-1-proficient strains are able to rupture the phagosomal membrane and obtain access to the host cell cytosol at later stages of the infection, whereas strains carrying a partial or interrupted ESX-1 system remain locked in the phagovacuole.[84,85] Hence, the ability of M. tuberculosis to break the phagosomal membrane and access the cytosol seems to be tightly linked to secretion ESAT-6 and CFP-10. The observed activity seems to be specific for ESAT-6 of M. tuberculosis[86,87] that, under acidic pH conditions, may undergo significant conformational changes, which is not the case for orthologous proteins from the nonpathogenic M. smegmatis species.[88] The ESX-1-dependent ability of inducing phagosomal breakage seems to be a key factor during the infection of host cells by M. tuberculosis, as only strains with intact ESX-1 secretion can cause cell death[85,89,90] and have enhanced cell-to-cell spread.[86,91] The cytosolic contact of ESX-1-proficient strains also determines the autophagic flux,[92,93] and may promote cross-presentation of mycobacterial antigens by the cytosolic MHC class I-processing machinery (i.e., the proteasome). The differences in cytosolic contact and induction of host cell necrosis between ESX-1-containing M. tuberculosis and naturally ESX-1-deleted BCG are also relevant for NLRP3 inflammasome activation, which leads to IL-1β secretion,[94–96] the generation of type I interferons[97] and impacts induction of CD8+ T-cell responses.[98] The identification of small-molecule inhibitors specifically targeting the ESX-1 secretion machinery or its secreted proteins may represent a suitable alternative way to neutralize the virulence factors of M. tuberculosis.[99,100] Without being bactericidal themselves, these virulence inhibitors could accentuate the innate and adaptive immune responses of the host on bacteria that have become devoid of their ESX-1-linked intracellular defense strategies. Such a strategy, which might work best in combination with conventional, bactericidal, anti-TB drugs, could also help with avoiding the generation of drug resistance due to the action of the immune system on potential escape mutants.

Inhibition of the ESX-1 system might also be achieved by targeting the regulation process of ESAT-6/CFP10 secretion, which is linked to the ESX-1-associated proteins EspA, EspC and EspD. The espACD locus, located in a different region of the M. tuberculosis genome than the ESX-1 core components, is important for ESX-1 functions because EspA, EspC and EspD are secreted in an ESAT-6/CFP-10-codependent fashion.[101,102] The expression of EspACD is regulated by a mechanism that involves the virulence regulator EspR,[103] which was recently described as a nucleoid-associated protein,[104] and the two-component regulators MprAB and PhoPR.[105,106] PhoPR is well known for its implication in the virulence of M. tuberculosis[107] and its large regulon that accounts for more than 40 positively and 70 negatively regulated genes.[108] Although the exact regulation cascade of PhoPR, MprAB and EspR in association with EspACD remains unclear,[109] the strong ESX-1-mediated implication of the espACD operon in the virulence of M. tuberculosis makes these regulators potential drug targets.

In analogy with the ESX-1 system, the ESX-5 secretion system might also represent a potential drug target, as it is simultaneously involved with the in vitro viability and in vivo replication of M. tuberculosis.[65,66,110] It was recently shown that disruption of certain ESX-5 core components, such as the predicted transmembrane channel protein EccD5 or the membrane-bound ATPase EccC5, affects the growth of M. tuberculosis on solid medium, yielding small colony morphotypes of mutant colonies in comparison with the wild-type strain.[64,65,110] ESX-5 was also shown to be essential for mycobacterial cell wall integrity, which is further confirmed by the enhanced detergent and hydrophilic antibiotic sensitivities of ESX-5-knockout strains.[65,110] The function of the ESX-5 system in M. tuberculosis appears to be mainly linked to the secretion of Esx and PE/PPE proteins, with the latter corresponding to two large protein families in M. tuberculosis that are named and classified after a characteristic Pro–Glu or Pro–Glu–Glu motif at the N-terminus class, as well as amino acid sequence similarity.[11,111] The functions of these immunogenic proteins[66,112] remain largely unknown, but it is clear now that some of them play a role in pathogenicity. For example, upon the deletion of the genomic section encoding PPE25-PE19 within the ESX-5 locus, the virulence of the resulting M. tuberculosis mutant was strongly diminished in immunocompetent and immunodeficient mice, whereas the complemented strain regained virulence.[65,66] Some of these proteins contain potent T-cell epitopes with or without cross-reactivity with other PE/PPE proteins[66] and might thus show good potential for inclusion into subunit vaccine combinations that might be used in combination with specific drug therapies as a form of immunotherapy, which is analogous to what has been tested in other bacterial species.[113]