Our previous work revealed that in steady state, CD44hi CD62Llo memory-phenotype (MP) CD4+ T lymphocytes are generated from peripheral naive precursors in a self-recognition-dependent manner. These cells contain a Th1-like T-bet hi subpopulation that produces IFN-gamma in response to pathogen-elicited IL-12 in the absence of antigen recognition, which augments antigen-specific CD4+ T cell responses and contributes significantly to host defense against Th1-inducing pathogens such as T. gondii, M. bovis BCG, and M. tuberculosis. (Kawabe T, et al., 2017). The above findings prompted us to address the mechanisms of how MP cells differentiate toward the innate-like T-bet hi MP subset under homeostatic conditions (i.e. with no foreign Ag stimulation). We found that IL-12 p40, an essential cytokine for Th1 differentiation, significantly promotes T-bet hi MP differentiation in steady state while it is dispensable for generation of T-bet lo MP and other types of innate lymphocytes including NK and NKT cells. IL-12 p40 is constitutively expressed by CD8 alpha + DCs with an activated phenotype (i.e., MHCII bright CD86hi CD40hi) and this DC subset plays a critical role in the optimal differentiation of T-bet hi MP cells. Furthermore, we revealed that this tonic IL-12 p40 production is induced by TLR MyD88 signaling and further enhanced by CD40L expressed by MP CD4+ T cells, in a manner independent of commensal antigen recognition. These finding suggests that in steady state T-bet hi MP cells are maintained as a consequence of IL-12 produced by CD8alpha+ DCs in response to an unknown endogenous stimulus distinct from that provided by the microbiota. This pathway is important in maintaining MP cells in a state that prepares them to respond innately to pathogen challenge. The host factors that determine whether individuals exposed to Mycobacterium tuberculosis (Mtb) become infected and progress to either latent infection or active disease are poorly defined. The microbiota has been identified as a key influence on the nutritional, metabolic, and immunological status of the host, although its role in the pathogenesis of Mtb is currently unclear. In studies performed this year, we investigated the possible impact of both the microbiome on the outcome of Mtb exposure as well as the effects of Mtb exposure on the microbiome. In the first project performed in collaboration with Dan Barber's research program, Rhesus macaques were infected via intrabronchial instillation with different strains and doses of Mtb in three independent experiments resulting in a range of disease severities. The composition of the microbiota was then assessed in fecal samples collected pre- and post-infection. Clustering analyses of the microbiota composition revealed that the microbiomes of monkeys that developed severe disease were very dissimilar from each other as well as from the microbiomes of the animals with less severe disease. In contrast, the microbiomes of the macaques with mild disease clustered together. These findings reveal that certain baseline microbiome communities are strongly associated with the development of severe tuberculosis following infection and may be more important as disease determinants than alterations induced in the microbiota due to Mtb infection itself. We extended the question of disease severity and correlation with microbiota composition to TB patients through a collaborative clinical study with our former post-doctoral fellow Dr.
Maiga aim ed at investigating the role of M. tuberculosis infection and treatment in Malian patients infected with either M. tuberculosis or M. africanum. M. africanum is a hypo-virulent mycobacteria that is localized to Western Africa and TB disease due to M. africanum is often observed in older and malnourished individuals as well as AIDS patients. Therefore, we wanted to evaluate if any differences in the intestinal microbiome exist between the two infections. Fecal samples were collected from 20 patients each with either M. tuberculosis or M. africanum at the time of diagnosis and 2 months following the start of anti-tuberculosis treatment. Fecal samples from 10 healthy donors were used as controls. Clustering analysis revealed that while mycobacterial infection by itself resulted in a small but significant difference in the microbiome structure, the two infected groups were not different from each other. Consistent with our previous findings (discussed further below), TB antibiotic treatment caused a significant alteration in the composition and structure of the microbiome in both infected groups. Interestingly , a comparison of the composition of the microbiome at the time of diagnosis between M. tuberculosis and M. africanum infected patients revealed a significant increase in the relative abundance of Proteobacteria, a phylum that consists of a large number of pathogenic bacterial species, in M. africanum patients in comparison with M. tuberculosis infected subjects This increase in Proteobacteria may reflect the more immune-compromised state of M. africanum patients or could be a consequence of their disease. Nevertheless, these findings demonstrate a preferential association of Proteobacteria with M. africanum versus M. tuberculosis induced tuberculosis that may be a factor in their distinct pathogenesis. As outlined in last years report we and our colleagues have demonstrated that treatment with conventional TB antibiotics results in a long-lasting dysbiosis in both experimentally infected mice and patients. Recent studies in other systems have linked changes in the microbiome with differences in drug metabolism and therapeutic outcomes. Therefore, in collaboration with Veronique DArtois (Rutgers Univ.) we investigated whether the microbiota might influence the absorption of anti-tuberculosis drugs and if the dysbiosis induced by treatment with these drugs affects their own absorption. To this end, we orally administered a combination of isoniazid, rifampin and pyrazinamide (HRZ) for 4 weeks to induce an intestinal dysbiosis. Additionally, a separate group of mice received a regimen for vancomycin, ampicillin, neomycin and metronidazole (VANM) for the same period to more drastically alter the microbiome. At the end of the treatment period, the mice in both antibiotic pre-treated groups and an untreated control group were orally administered one or a combination of anti-TB drugs. Blood samples were collected over a 12 hr period post drug administration and plasma levels of the administered antibiotics determined. In comparison to mice with an undisturbed microbiome, animals with an HRZ-induced dysbiosis actually showed increased rather than decreased circulating levels of rifampicin and moxifloxacin when ravaged with these antibiotics. Interestingly, mice pretreated with VANM showed the opposite effect with significantly lower levels of both antibiotics in comparison to the untreated controls as well as dysbiotic HRZ pre-treated animals. In contrast, we failed to detect a difference in drug plasma levels between HRZ or VNM pre-treated mice given pyrazinamide, an antibiotic that is absorbed by passive diffusion. The latter findings suggest that a dysbiotic microbiome preferentially affects the absorption of TB antibiotics that must be first metabolized.
Showing the most recent 10 out of 30 publications
