In this project, we infected mice with the M. bovis vaccine strain BCG Pasteur to model Mtb infection. We collected serum, lung and bronchoalveolar lavage samples over time. We were able to follow the induction of antibody responses (isotypes IgG, IgA and IgM) against BCG over time. When antibodies in serum from mice that were infected with BCG were transferred into naïve mice prior to challenge with BCG, the bacterial load in the lung of infected mice was lower compared to mice that had not received serum prior to BCG challenge. This suggests that antibodies might help to protect against infection with mycobacteria in vivo. BCG-induced antibodies targeted surface-epitopes of BCG as well as two clinical strains of M. tuberculosis. Because such surface-binding antibodies often contribute towards protection against infection with bacteria, we developed a procedure to extract surface-exposed proteins of BCG and determined the identity using mass spectrometry. The results identified a subset of both already known antigens as well as proteins that might present interesting antibody targets in future. To further clarify what bacterial proteins antibodies bind to, we successfully developed an assay that uses immunoproteomics as an unbiased high-throughput approach to identify which antigens are targeted by antibodies after infection of mice. We applied this approach to a cohort of mice that received BCG via intravenous (IV), intratracheal (IT) or subcutaneous (SC) route. Antibodies against the surface as well as intracellular proteins of BCG were greatly induced after IV as well as IT BCG administration, whereas SC BCG administration only led to a modest increase in IgG antibody levels in serum, bronchoalveolar lavage and lung. Using our immunoproteomic assay, we identified proteins that were targeted by antibody responses for the different vaccine routes and tissues. These candidates will be further validated and might serve as interesting vaccine targets in future.
To link antibody specificity with antibody function, our aim was to isolate individual B cells and sequence the antibodies they encode. While this work is still in progress, we already developed in vitro infection assays that can be used to test whether a monoclonal antibody is protective against infection with mycobacteria. Here, we created fluorescent reporter strains of BCG that enable the quantification of how many bacteria are taken up by different macrophage cell population – the cell type that is a major target for Mtb at early stages of infection – and give information about the intracellular fate of mycobacteria that were taken up by macrophages. Being able to link antibody specificity with antibody function will ultimately enable to identify antibodies that can protect against infection with Mtb. Such antibodies will represent valuable research tool for the TB research community.
Lastly, we aimed to translate our findings using the mouse model back into humans. Here, the project included the use of samples from a clinical study of TB patients with different grades of TB susceptibility. By characterising what mucosal IgG antibodies bind to, we observed differences between the systemic antibody response in serum and the lung antibody response, a focus of the antibody response towards protein and polysaccharide antigens as well as an increase of antibody response correlating with the degree of exposure patients towards Mtb infection.
