Pathway biology of the systemic host response in human tuberculosis- a systems meta-analysis

Tuberculosis caused by members of the Mycobacterium tuberculosis complex is a leading cause of morbidity and mortality due to infectious disease. Infection with M. tuberculosis results in recruitment of immune cells into the lungs where granulomata are formed in which the bacteria have limited nutrient and carbon sources with the exception of lipids. There is increasing evidence linking immunity and metabolism; including metabolites generated in the cholesterol biosynthesis pathways, in particular, oxidised cholesterol (oxysterols). Two immune-regulated oxysterols, 25 hydroxycholesterol (25HC) and vitamin D metabolites (1,25 dihydroxyvitamin D2 and 1,25 dihydroxyvitamin D3) are also known to have strong physiological effects on immune responses. Work in this thesis investigates host immune responses and links to lipid metabolism to tuberculosis infection using hypothesis and data driven approaches. The effects of 25HC and vitamin D metabolites in mycobacterial growth were investigated in 7H9 culture media in vitro and in whole blood ex vivo using a growth inhibition assay (BCG-lux). These studies demonstrated that supplementation of 7H9 with these compounds does not directly affect the mycobacterial growth. Furthermore, whole blood mycobacterial growth inhibition assays indicate lack of effect on mycobacterial growth. These results suggest, under the conditions tested that host oxysterols do not have a direct effect on mycobacterial growth cycle. However, the question of whether sterols have a role in vivo remains. To gain insight into in vivo host systemic host responses to tuberculosis, a metaanalysis of five adult tuberculosis (TB) datasets from 4 different study sites consisting of 692 tuberculosis patients and controls and two childhood TB datasets from 3 different study sites consisting of 257 subjects was conducted. Results from the meta-analysis revealed statistically significant modulation of genes involved in immunity and cellular metabolism. Specifically, there was up regulation of innate immune response genes such as defensins, MPO, C1Q components, TLRs including TLR2, 4 and 5, interferon inducible genes such as IFITM, IFI44, and IFITM in tuberculosis patients. Adaptive immune response genes including T cell signaling genes such as CD40, ICOS, ITK and ZAP70 were down regulated in active tuberculosis patients. Metabolic genes including GPR84, ALPL, RETN, SLC2A3 and GYG1 recently identified as part of a neonatal sepsis classifier, were all up regulated in active tuberculosis. This data suggests significant modulation of host immune metabolic responses in tuberculosis. Next, differentially expressed genes from the meta-analysis were used to investigate in vivo systemic modulation of the sterol biosynthesis pathway. Results obtained suggested accumulation of metabolites in the cholesterol biosynthesis pathway in active tuberculosis due to up regulation of isopentenyl diphosphate isomerase (IDI) and down regulation of lanosterol synthase (LSS). Accumulation of intermediate metabolites such as dimethylallyl pyrophosphate (DMAPP) could potentially affect the host immune response as stimulators of natural killer cells and also as potential carbon sources for mycobacterial anabolic pathways. Genes involved in cholesterol transport including APOL1, APOL2 and ABCA1 were up regulated in active tuberculosis patients. This suggests a potential role of host sterol metabolism in tuberculosis pathogenesis and immunity. Finally differentially regulated genes were subjected to global pathway analyses to determine systemic host modulated pathways. Results suggested up regulation of innate immune pathways including TLR and NOD-like receptor signaling pathways, which lead to the induction of pro-IL-1β and its conversion into mature IL-1β by the inflammasome and the expression of IP-10. Also up regulated were the complement and coagulation and systemic lupus erythematosus pathways. In contrast, adaptive immune pathways were supressed in active tuberculosis including T cell signaling due to down regulation of proximal signaling molecules including CD3 components CD28, LAT and ITK. Cytokine-cytokine receptor interaction and B cell receptor-signaling pathways were also down regulated in active tuberculosis. Furthermore, in silico white blood cell deconvolution analysis based on the gene expression signals indicated higher and lower proportions of innate and adaptive immune cells respectively in active tuberculosis compared to latently infected and uninfected healthy controls. In summary these investigations suggest modulation of host immune metabolic pathways in tuberculosis. This could have an important implication in the effort to enhance the immune response to mycobacterial infection through the modulation of the immune-metabolic-axis in host-directed therapy