Unit: Mycobacterial Genetics
Director: GICQUEL Brigitte
Tuberculosis remains a major public health problem responsible for about two million deaths each year. It is caused by mycobacteria from the Tuberculosis complex.The only currently available vaccine is BCG, which is relatively efficient. The efficacy of antibiotics for treating tuberculosis may be reduced by the emergence of multiresistant strains of bacteria. Our unit characterises Mycobacterium tuberculosis virulence factors and the host immune responses induced by this bacterium and BCG. Our work may help identify targets for new drugs against tuberculosis, help identify bacterial components that could be included in a new vaccine, or may even lead to a new attenuated strain more efficient than BCG. Our Unit collaborates with other teams to identify host genetic markers associated to the susceptibility to tuberculosis. We also search for genetic markers specific to epidemic M. tuberculosis strains, particularly strains responsible for multidrug-resistant outbreaks.
Genetic study of virulence determinants (M. Jackson, O. Neyrolles, B. Gicquel)
We used a temperature-sensitive vector derived from the pAL5000 plasmid of Mycobacterium fortuitum and containing the counterselective marker sacB (Ts/sacB vector) to construct allelic exchange and transposon mutants in a mycobacterial species of the M. tuberculosis complex. We constructed an ordered transposition mutant library of M. tuberculosis containing 4000 mutants using signature-tagged transposon mutagenesis (STM) and directly screened for mutants exhibiting an attenuated virulence in mice. About 4000 mutants have now been screened for their ability to multiply in the lungs of mice during the acute phase of infection. We selected 31 mutants and characterised the corresponding mutations.
We identified ten different transposon insertions in a 50 kb-region of the genome dedicated to the synthesis and transport of phthiocerol dimycocerosates (DIM) and phenolic glycolipids (PGL). Several genes from this region were functionally characterised. We have recently shown that lppX encodes a lipoprotein involved in translocating DIM to the cell surface of M. tuberculosis. The three-dimensional structure of LppX (determined in collaboration with the laboratory of Dr. Y. Bourne at the CNRS in Marseille) revealed that LppX has a similar folding pattern to LolA and LolB (Sulzenbacher et al., EMBO J., 2006). In Gram negative bacteria, LolA and LolB translocate lipoproteins from the periplasm to the outer membrane. This is the first proposed mechanism for the transport of lipids to the outer membrane of mycobacteria. This work also highlights the important role that lipoproteins play in building the mycobacterial cell envelope.
We also biochemically characterised other M. tuberculosis transposon mutants with deficiencies in the synthesis of the saccharidic and phenolic moieties of PGL and we are currently evaluating their residual virulence in macrophage and mouse models of infection.
Parallel with this, we have also studied the M. tuberculosis two-component signal transduction system PhoP-PhoR. This system has sequence similarities with the PhoP-PhoR regulator of Salmonella, which is known to regulate more than forty genes, among which are several key virulence factors. In collaboration with Prof. Carlos Martín's group (Zaragoza Medical School, Spain), we showed that PhoP-PhoR is required for the replication of the tubercle bacillus in cultured macrophages and mice. We then aimed to identify the genes regulated by PhoP-PhoR to define the molecular determinants underlying the attenuation of the virulence of the phoP mutant. As the phoP mutant exhibited altered colonial morphology and cording properties, we first compared the lipid content of phoP and phoP-phoR mutants constructed in two different strains of M. tuberculosis to that of their wild-type parent. Our results showed that PhoP co-ordinately and positively regulates the synthesis of polyketide-derived acyltrehaloses, namely diacyltrehaloses, polyacyltrehaloses and sulfolipids, that are known to be restricted to pathogenic species of the M. tuberculosis complex (Asensio et al., J. Biol. Chem, 2006).
The protective efficacy of the attenuated M. tuberculosis strains isolated in the laboratory is being studied as part of the EU FP6 program "TB-VAC".
Interactions between M. tuberculosis and phagocytes (O. Neyrolles, M. Jackson, B. Gicquel)
We are currently developing new approaches combining cell biology and functional genomics to understand better the mechanisms of macrophage parasitism by M. tuberculosis. We also used STM to identify M. tuberculosis genes involved in the ability of the bacillus to parasitise human macrophages. We have isolated 21 attenuated mutants and have identified the carried mutations. The identified genes are currently being studied.
We are also studying early interactions between mycobacteria and human phagocytes. We have recently shown that the lectin DC-SIGN is the major M. tuberculosis receptor on human dendritic cells, and that mycobacterial lipoarabinomannan (LAM) is a key DC-SIGN ligand. We have also detected DC-SIGN expression on human lung dendritic cells and have detected mycobacteria in DC-SIGN-positive dendritic cells in lymph node biopsies from patients with tuberculosis. These results suggest that DC-SIGN-mediated interactions between M. tuberculosis and dendritic cells should occur in vivo. We have also identified the molecular determinants of the LAM-DC-SIGN ligation. These explain why DC-SIGN specifically recognises pathogenic mycobacteria from the "tuberculosis" complex. Altogether, our results show that pathogenic mycobacteria may have evolved surface motifs to interact with lectins, such as DC-SIGN, on the surface of phagocytes. We have also shown that the selective recognition of M. tuberculosis by DC-SIGN most probabaly relies on multiple ligand recognition by the lectin within the mycobacterial envelope. (photo 1)
Furthermore, in collaboration with Saint-Louis and Necker Hospitals in Paris, we have observed DC-SIGN induction in alveolar macrophages from patients with TB, but not in cells from patients with other lung diseases or from control subjects. We have shown that DC-SIGN-expressing macrophages are a preferential host cell population for mycobacteria in patients with TB (Tailleux et al., PloS Medicine, 2006). The functional consequences of DC-SIGN ligation by the bacillus in the lungs of infected individuals are being investigated. Moreover, a case-control study in which our team has participated has shown that in a South-African cohort, a variant of the promoter of the gene encoding DC-SIGN is strongly associated with protection against TB (Barreiro et al., PloS Medicine, 2006). This result further highlights the important role played by the receptor during TB infection.
Identification of new anti-TB drug targets (M. Jackson, B. Gicquel)
Phosphatidylinositol (PI) and metabolically-derived products such as the phosphatidylinositol mannosides (PIM), the linear and mature branched lipomannan and lipoarabinomannan are prominent phospholipids/lipoglycans of Mycobacterium spp. that are thought to play important roles in the physiology of the bacterium and during host infection. As well as providing fundamental knowledge about the synthesis of these molecules, the characterisation of the enzymes involved in their biosynthesis may help identify attractive drug targets for the development of new anti-TB drugs.
We have identified a cluster of five genes that are potentially dedicated to the early steps of PIM synthesis. We examined the function of two of these genes, pimA (Rv2610c) and Rv2611c, and showed that they respectively encode a mannosyltransferase and an acyltransferase involved in the synthesis of di- and tri-acylated phosphatidylinositol mono-mannosides. We have also shown that PimA is essential for mycobacterial growth. Therefore, PimA may be an attractive novel drug target. The Structural Biochemistry Unit of the Institut Pasteur recently solved the structure of the PimA protein from M. smegmatis. Knowledge of the 3D-structure of this protein will allow potential inhibitors to be screened by computer analysis. The inhibitory activity of the identified compounds will then be tested in vitro using the cell-free assays that we have developed in the laboratory.
A broader approach was to identify other M. tuberculosis enzymes involved in the biosynthesis of PIM, LM and LAM by knocking out several glycosyltransferase genes potentially involved in the synthesis of these molecules by homologous recombination.
We are also studying, as part of our search for novel therapeutic targets, the mode of action of Isoxyl, an old anti-TB drug that was used for the clinical treatment of tuberculosis in the 1960s. Isoxyl inhibits oleic acid and mycolic acid synthesis in mycobacteria.
BCG and new vaccines (N. Winter)
Mycobacterium bovis BCG is the only available form of prophylaxis for controlling tuberculosis. Although BCG use is widespread, the immune mechanisms leading to partially effective vaccination are poorly understood. The innate immune response, which occurs very early after the introduction of the vaccine, is probably a decisive factor in the ensuing protective response. Thus, we injected BCG in the ear dorsum of mice as a surrogate of intradermal vaccination in humans. During the first three days, we tracked BCG host cells migrating out of the dermis to the auricular draining lymph nodes. Resident skin DCs or macrophages did not play a major role in early BCG capture and transport to the lymph nodes. Neutrophils were the principal BCG host cells rapidly recruited in both the dermis and the lymph nodes. Fluorescent green or red BCG strains injected into nonoverlapping sites were essentially sheltered by distinct neutrophils in the ADLN capsule, indicating that neutrophils had captured bacilli in peripheral tissue and transported them to the lymphoid organ. We observed by confocal microscopy on ear dermis, BCG-infected neutrophils in the lumen of lymphatic vessels. Fluorescent labelled neutrophils injected into the ears accumulated exclusively in the ipsilateral ADLN capsule after BCG vaccination. Thus, we provide in vivo evidence that neutrophils, like DCs or inflammatory monocytes, migrate via afferent lymphatics to lymphoid tissue and can shuttle live microorganisms (Abadie et al., Blood, 2005). In the lymph node, we observed neutrophils close to DCs and T-cells, suggesting that they may play a role in antigen presentation. We are currently investigating if this is the case. We are also deciphering the molecular mechanisms that allow activated neutrophils to leave the tissue and migrate to the lymph node.
After intravenous inoculation of BCG, we observed that DCs could host live BCG in vivo. Therefore, we would like to investigate how directly infected DCs process and present BCG-borne antigens to naive T-lymphocytes. We are currently investigating how the interplay between signals delivered through Toll-Like Receptors 2 and 4 and the antigen processing pathways in directly infected DCS restrict MHC class I and II presentation to naive CD8 and CD4 T-cells. (photo 2)
Molecular Epidemiology of Tuberculosis (Brigitte Gicquel)/p>
We looked for genetic markers specific for highly transmissible M. tuberculosis strains responsible for outbreaks, and in particular, multi-drug resistant (MDR) outbreaks. This study was carried out in collaboration with the European Network of Molecular Epidemiology, the Public Health Research Institute and the network of Pasteur Institutes and associated Institutes. We have identified specific alleles of putative genes involved in DNA repair. This suggests that antimutator genes may play a role in the evolution of several M. tuberculosis family strains. We found mut T- and ogt-specific alleles in M. tuberculosis strains of the W-Beijing genotype. These variations, which are specific to M. tuberculosis genotypes, may be useful for identifying different branches of the W. Beijing family through molecular diagnostics.
The four mutT annotated genes were investigated. We found that mutT1 is an antimutator in M. tuberculosis and M. smegmatis. However, mutT4 was found to be responsible for an antimutator phenotype only in M. smegmatis (Dos Vultos et al., J. Bacteriol., 2006).
We are using molecular methods to type M. tuberculosis strains from different geographical areas. New specific polymorphic genetic markers of major family strains have been identified. This should allow us to identify antimutator genes and the major family strains by genotyping techniques and allow their transmission to be studied.
Photo 1 : DC-SIGN expression at the surface of a cell.
Photo 2 : Mouse bone-Marrow-derived DCs infected with a rBCG strain expressing a green fluorescent protein. H2M molecules the invariant chain (red) concentrate in vacuoles surrounding BCG bacilli (green) whereas MHC class II molecules (blue) migrate to the surface of the cells indicating maturation. (Confocal microscope picture, courtesy of Dr Eric Prina)
Keywords: Mycobacterium, pathogenicity, virulence, tuberculosis, epidemiology, vaccine