Lung tissue resident memory T and B cells in the immune response to Mycobacterium tuberculosis in humans.

Abstract

Tuberculosis (TB), caused by Mycobacterium tuberculosis (Mtb) is the main cause of death from a single infectious agent. Although a quarter of the world population is estimated to be infected with TB, only 5–10% of these individuals will develop active disease during their lifetime, suggesting that the immune system is highly effective at containing Mtb infection in most people. Extensive analysis of T-cells in blood has, however, failed to identify correlates of protection in humans, which is a prerequisite for rationale vaccine design. A likely explanation for this failure is that most memory T-cells in lung tissue do not recirculate in blood. Indeed, adoptive transfer of lung T-cells from infected mice protects recipients from Mtb challenge whereas transfer of blood T-cells does not. While the crucial role of T-cells in controlling Mtb infection is well appreciated, the role of B cells and humoral immunity in TB disease is unclear. Furthermore, human studies that directly compare the B-cell involvement in TB between infected lung tissue and matched peripheral blood are lacking. Therefore, this study was based on the premise that studying the human adaptive immune responses to Mtb at the site of infection would provide new findings that are not present in the systemic circulation. Thus, we hypothesised that the lung T and B-cell responses to TB in humans are fundamentally different from those in the blood, because they involve cells that do not recirculate (particularly T-cells). The overall aim was to determine the phenotype, function and TB specificity of adaptive immune responses (T and B-cells) at the site of TB disease and to understand how they correlate with the responses detectable in circulation. To achieve this, a cohort of study participants undergoing medical lung resection for tuberculosis related lung pathologies was established The results presented in this thesis show that lung T-cells are different from peripheral blood T-cells in terms of phenotype, function, and specificity. Lung tissue contains a distinct population of effector memory T-cells expressing canonical markers of Tissue Resident Memory (Trm) T-cells, including CD69 and CD103, which are expanded during active TB disease. These Trm are highly functional, are enriched for IL-17 and IFN-γ secreting cells, and are depleted by HIV co-infection, compared to matched blood samples. In addition, they are highly enriched for TB-specific T-cells, including TB-specific Th17 and regulatory T-cell subsets. Th17 TB-specific T-cells in the lung negatively correlate with plasma cytokine markers of inflammation, suggesting an important role for this subset in human lung in limiting disease severity. Interestingly, although CD103 expressing T-cells produce the highest levels of cytokine by non-specific stimulation, TB-specific T-cells in the lung lack CD103 expression. Finally, detailed phenotyping of lung Trms was conducted using Mass cytometry and a panel of 40 T-cell markers. This confirmed patterns of expression of markers that are consistent with published data, including CXCR3, PD1, KLRG1 and other integrins. The existence of Trms in TB infected lung tissue was also investigated by global TCR sequencing. This revealed a significant clonal expansion of T-cells in the lung tissue compared to peripheral blood, consistent with a non-recirculating tissue resident memory population. When T-cells from blood and lung tissue were pre-sorted by flow cytometry before TCR sequencing, the results clearly demonstrated that the lung tissue is made of up to 20% clonally expanded T-cell clones that do not recirculate. This approach was also used to investigate the potential contribution of non-classical, donor unrestricted T-cells to lung immunity. By flow cytometry and TCR sequencing, we show that unconventional T-cells such as mucosal associated invariant T-cells (MAIT), germline-encoded mycolyl-reactive (GEM) T-cells and γδ T-cells are all depleted in peripheral blood of TB infected individuals compared to healthy controls while iNKT cells are not. However, this is unlikely to be due to recruitment to the site of disease, as no subset was consistently expanded in the lung, GEMs were rarely detected and γδ T-cells were present at a much lower frequency than in matched blood. Indeed, this study demonstrates the presence of unique and expanded γδ T-cells in TB infected lung that are absent in circulation, suggesting some DURTs may also exist as nonrecirculating lung resident memory cells. Analysis of B-cells in individuals with TB revealed skewing of blood B-cell phenotypes in both active and latent TB infection and an enrichment of total B-cells in the lung compared to matched blood. Lung B-cells were mainly of memory B-cell phenotype including plasmablasts, and, like T-cells, were enriched for the expression of the Trm marker CD69, suggesting that B-cells in the lung probably do not recirculate. This potential Trm B-cell population was enriched for B-cell activation markers, CD95 and HLA-DR, suggesting involvement at the site of disease. However, using TB-specific BAITs, the frequency of TB specific B-cells in the lung tissue was found to be lower compared to TB specific B-cells in blood. The functional role of B-cells in TB disease warrants further investigation. These results provide clear evidence of compartmentalization of the adaptive immune responses to Mtb in the human lung, which maybe important when interpreting data from human PBMC for correlates of protection or vaccine response.