M. tuberculosis lineages: genetic diversity and its involvement on macrophage infection and on drug tolerance

A central feature of Mycobacterium tuberculosis pathogenesis is the ability to survive within macrophages and colonize hostile environments that are acidic and rich in cholesterol and fatty acids. The genetic variability among clinical isolates may have dramatic consequences on the outcome of infections. Many in vitro and in vivo studies have demonstrated strain-dependent variation in key aspects of virulence such as stress survival, transmission, pathology, and lethality. This variability in MTB clinical isolates went neglected for decades; however, recently the scientific community recognized it causes important consequences on the progression of infection. As reduced host response is supposed to be the key to enabling MTB persistence and transmissibility, the immunological state of the host is also important to consider when studying the relative virulence of clinical MTB strains. In this context, the polarization states of macrophages have a huge impact on their function. It affects how macrophages can react to external signals, changing gene expression, membrane composition, receptor exposure, and cytokines production. In our work, we focused our attention on MTB diversity and its role in pathogenesis and drug resistance acquisition. We proposed a comprehensive approach for better understanding MTB virulence taking into consideration both human (phenotypic) and MTB (genetic) variability. We selected well-characterized strains belonging to specific MTB lineages and adopt the THP-I-derived macrophage infection model, but considering host variability by deriving M1 or M2 polarized macrophages. Confocal live microscopy was used to study at a single-cell level different macrophage pathways during the infection. These included phagolysosomal acidification, autophagic flux, and apoptosis. All are mechanisms that MTB hijacks in order to survive within the host. Cytokine profiles were measured in response to the different MTB lineages, along with the survival of both the macrophages and the bacteria at different time points post-infection. In M1 macrophages, whereas less virulent/ancient are more efficient in blocking the phagosomal acidification, more virulent/modern strains are likely better at tolerating the acidic environment. However, the number of responsive macrophages remained relatively low. In M2a macrophages, whereas less virulent/ancient strains are less efficient in blocking the phagolysosomal acidification, more virulent/modern strains are blocking phagolysosomal acidification. The number of responsive macrophages was found to vary depending upon the features of the infecting strains, with the most virulent ones inducing the lowest percentages of acidifying cells. The analysis of the autophagic flux showed less heterogeneity among both the bacterial strains and the macrophage phenotypes considered. Indeed, the induction of autophagy was found negligible in our observations. In M1 macrophages, bystander non-infected cells of more virulent/ modern showed increased apoptosis (40%). Stratification of the data by mycobacterial burden showed similar apoptotic levels irrespective of the mycobacterial load, despite a trend displaying slightly higher apoptotic levels in bystander non-infected cells or in cells infected with a lower number of mycobacteria could be noted. In contrast, none of the categories induced apoptosis in M2 macrophages at the time point considered. In M1 cells, the infection with less virulent/ancient lineages reduced the production of pro-inflammatory IL1-β. Likewise, IL-18 levels were lower in M1 cells infected with less virulent/ancient lineages. The chemokine GROα showed a similar pattern between MTB categories considered. In M2 cells we observed the modulation of anti-inflammatory IL-10 when comparing clinical to laboratory strains. Most of the statistically significant differences found were observed within 24h for both polarization types. Our findings on macrophage survival corroborate the results obtained at the single-cell level, with more virulent/modern lineages causing increased cell death. Interestingly, these data seem to contrast the finding that such strains showed lower macrophage entry in colony-forming units evaluation studies. Similarly, there was not a direct correlation between pro-inflammatory cytokine production and macrophage killing. Nevertheless, more virulent/modern strains showed increased intramacrophagic replicative capacity compared to less virulent/ancient lineages. We also studied the genetic variability of the MTB strains during drug resistance acquisition. Among newly investigated mechanisms, the role of smallRNAs in the development of drug resistance has been considered in other bacteria but remains unexplored in MTB. We characterized the smallRNA ncRv0842c, cis-encoded to the Rv0842 gene which codes for a putative efflux pump reported to be involved in rifampicin resistance (RIF-R) development. Accordingly, the present study characterized the role of ncRv0842c during RIF challenge and validated the role of a lineage-specific synonymous mutation on the coding region of Rv0842 affecting the promoter region of the cis-encoded smallRNA. In order to understand the role of the smallRNA in modulating the efflux pump, we characterized its expression during challenge experiments using sub-inhibitory concentrations of rifampicin. We then generated overexpressing mutants from MTB-selected ancient and modern strains. qPCR showed basal downregulation of ncRv0842 in ancient lineages compared to modern lineages. This result confirmed the hypothesis that the synonymous mutation in Rv0842 (L45L) specific for the ancient lineages, and abrogating the -10 promoter region of the antisense smallRNA is severely affecting the expression of the smallRNA. Under RIF-induced stress, we observed upregulation of Rv0842, and downregulation of the smallRNA only in H37Rv (modern lineage) whereas in L5 (ancient lineage) the smallRNA expression was not affected. MABA assay performed on modern lineage mutants overexpressing ncRv0842c showed a 1-dilution reduction of the MIC in comparison with their respective control (mock) while the ancient lineage strains did not show any MIC shift despite the overexpression of the smallRNA. Our analysis showed that the unraveling of smallRNA may provide new insights on MTB lineage-specific adaptation to drug-related stress and uncover the role of silent mutations in determining different phenotypes in MTB. Our study can contribute to better understanding lineage-specific pathogenicity and highlights the importance of strain- and lineage-specific rational design and development of effective diagnostic tools and vaccines.