Infection Immunology

Protective granulomas can suppress an infection with Mycobacterium tuberculosis over the long term. They are functionally organised and contain long-lived immune cells that enable bacterial control. This form of protective immune organisation is primarily supported by pro-inflammatory T helper cells (Th) of the Th1 and Th17 types, which provide specific signals to activate infected macrophages and enhance antibacterial mechanisms in the lungs. At the same time, their activity is regulated by inhibitory cytokines to limit tissue damage. In particular, the pro-inflammatory cytokine interleukin (IL)-17, which is mainly produced by Th17 cells and gamma-delta T cells, plays an important role during granuloma development. In order to gain a more detailed understanding of the cytokine-mediated control mechanisms for establishing and maintaining protective granulomas, the Infection Immunology research group is interested in the differential roles of cytokines and cytokine receptors associated with the ubiquitous receptor subunit gp130. In this context, we were able to show that the gp130-dependent cytokine IL-6 plays a minor role in the development of IL-17A-expressing Th17 cells during experimental tuberculosis (TB), in contrast to its key role in the induction of Th17 immune responses in other inflammatory diseases. Accordingly, the absence of IL-6 or gp130 on T cells has little effect on the containment of the infection. In contrast, a global deficiency of the cytokine receptor subunit IL-27Rα in M. tuberculosis-infected mice leads to reduced bacterial load, but also to increased immunopathology. IL-27Rα together with gp130 forms the receptor of the heterodimeric cytokine IL-27. The enhanced containment of mycobacteria in IL-27Rα-deficient mice is accompanied by the formation of highly structured protective granulomas, which consist of a core of infected macrophages surrounded by a rim of protective T and B cells. By analysing the course of M. tuberculosis infection in IL-27Rα x IL-17A double-deficient mice, we were able to further demonstrate that the protective effect of IL-27Rα deficiency is dependent on the correlating increased expression of the cytokine IL-17A. IL-27 thus appears to limit the development of a protective immune response against M. tuberculosis, partly through the inhibition of IL-17A. At the same time, however, the suppression of IL-17A expression by IL-27 also leads to a reduction in immunopathological consequences. Most recently, we were able to demonstrate in mice with a cell type-specific deficiency of IL-27Rα-mediated signalling in regulatory T cells that the inhibitory effect of IL-27 on the development of protective granulomas is partly due to direct IL-27-mediated induction of these immunosuppressive cells (Fig. 1). Overall, this work provides crucial foundations for the further development of adjuvant therapeutic strategies with the aim of optimising the architecture of granulomatous lung lesions through targeted immunomodulation.

Fig. 1. Cell type-specific deficiency of IL-27Rα-mediated signalling in regulatory T cells (T_reg) promotes the formation of highly structured protective granulomas. (A) Mice with a T_reg -specific IL-27Rα deficiency (middle column) form structured granulomas in the lungs after infection with M. tuberculosis, similar to mice with global IL-27Rα deficiency (right column). These granulomas consist of a lymphocyte ring (top row) and a central core of activated macrophages (bottom row). (B) In the chronic phase of M. tuberculosis infection, the bacterial load in T_reg -specific IL-27Rα-deficient mice (medium grey bars) is reduced compared to the control group (light grey bars).

Damaging granulomas are caused by a misdirected immune response in which protective mechanisms for controlling pathogens are converted into tissue-destructive processes. The central histopathological feature of human TB, central granuloma necrosis following infection with M. tuberculosis, does not typically occur in standard laboratory mice. Based on our investigations in TB patients, the Infection Immunology research group described a functional SNP in the IL4RA gene that is associated with enhanced signal transduction and more severe pulmonary pathology. This led us to the hypothesise that a Th2-dominated immune response mediated by IL-4Rα contributes significantly to TB pathology. Since conventional mice do not develop a corresponding Th2 immune response after infection with M. tuberculosis, we infected IL-13-overexpressing mice with M. tuberculosis. In these animals, the IL-13-driven Th2 immune response via the IL-4Rα signalling pathway led to the formation of central necrotising granulomas, which are the essential histopathological feature of human post-primary TB. This provides the Infection Immunology research group with an experimental model that enables the targeted analysis of downstream immunological and metabolic mechanisms of TB-associated tissue pathology. In this context, there is pronounced alternative activation of macrophages with strong arginase-1 expression, lipid accumulation and formation of a hypoxic necrotic granuloma core surrounded by a collagen capsule. These arginase-1-positive foam macrophages promote both pathological tissue remodelling and bacterial persistence within the granuloma. In addition, we were able to show that damaging granuloma necrosis can also be induced independently of IL-13 when the NOS2-dependent antimicrobial metabolic pathway is pharmacologically blocked by the inhibitor L-NIL. In standard C57BL/6 mice, this targeted inhibition leads to a shift in L-arginine metabolism towards arginase-1 activity and the formation of centrally necrotic granulomas with fibrotic collagen rings. Crucially, this pathology is completely eliminated in mice with macrophage-specific knockout of arginase-1, even though NOS2 inhibition remains intact (Fig. 2). The work of the Infection Immunology research group thus demonstrates that arginase-1-dependent mechanisms are a central, common endpoint of damaging granuloma formation and contribute significantly to the pathology, resistance to therapy and progression of TB.

Fig. 2. The absence of arginase-1 leads to improved defence against M. tuberculosis and modulated lung pathology. Administration of the NOS2 inhibitor L-NIL induces arginase-1 activity in M. tuberculosis-infected control mice (Tie2cre^neg  x Arg1^loxP/loxP). This activity is abolished in Arg-1-deficient animals (Tie2cre^pos  x Arg1^loxP/loxP ). (A) At various time points during infection, the lungs of the animals were removed and the bacterial load (CFU) was determined. (B) After 101 days, microscopic sections were made from the lungs and stained with haematoxylin/eosin (upper row) or immunohistologically with an anti-Arg-1 antibody (lower row).

Unlike standard inbred strains such as C57BL/6, both IL-13tg mice and C3HeJ/FeB mice develop the granulomatous pathology typical of human TB, with central necrosis, fibrosis and hypoxic microenvironments (Fig. 3).  Both models thus offer the opportunity to identify common pathogenetic features of these pronounced lesions and, on this basis, to uncover previously unknown downstream mechanisms of TB pathology and to evaluate the efficacy of antibiotics under pathophysiologically relevant conditions in a preclinical setting.

Fig. 3. In contrast to M. tuberculosis-infected BALB/c mice, IL-13tg and C3HeB/FeJ mice form centrally necrotising granulomas, similar to TB patients. (A) BALB/c, (B) IL-13tg and (C) C3HeB/FeJ mice were infected with M. tuberculosis via aerosol. After approximately 10 weeks, the lungs were removed, microscopic sections were prepared and the collagen was stained (blue areas). 

Antibiotic-resistant strains of M. tuberculosis are a major threat to global control of TB, which is why new anti-TB drugs are urgently needed. Drug development against multidrug-resistant TB (MDR-TB) requires preclinical validation in advanced animal models that reflect the pathophysiological aspects of human TB, since the complex structure of centrally necrotizing granulomas favours mycobacterial survival but hampers the penetration of many antibiotics. Therefore, the Infection Immunology research group, in collaboration with other scientists within the German Centre for Infection Research (DZIF), has established a powerful preclinical infrastructure for the improved evaluation of novel anti-mycobacterial compounds, which is also integrated into international networks (ERA4TB and UNITE4TB). A fundamental pillar of this preclinical platform are advanced mouse models (IL-13tg, C3HeB/FeJ), that develop a human-like granuloma pathology with a mycobacteria harbouring necrotic core and fibrotic encapsulation. For a comprehensive in vivo validation of novel drug candidates, we combine these advanced mouse models with state-of-the-art methods for drug detection, such as mass spectrometry imaging (MSI, collaboration with Andreas Römpp, University of Bayreuth) or laser capture microdissection (LCM) coupled with LC-MS/MS (collaboration with Dominik Schwudke, Research Centre Borstel). By combining LCM with molecular methods for the detection of M. tuberculosis (molecular bacterial load Assay, MBLA) we determine the pathogen burden within necrotic granulomas (Fig. 4). This innovative approach was used for the comprehensive validation of the novel antibiotic BTZ-043 revealing that BTZ-043 fully penetrates centrally necrotising granulomas, at concentrations manifold above its in vitro minimum inhibitory concentration (1 ng/ml) and that BTZ-043 exerts its antibacterial activity within necrotic lesions(Fig. 5). Based on these properties, BTZ-043 has the potential to complement existing TB antibiotics and may contribute to future regimens to shorten treatment duration, prevent reactivation and avoid the development of drug resistance. The efficacy of BTZ-043-based combination therapies are currently being evaluated in clinical trials (UNITE4TB and PanACEA).

Fig. 4. Preclinical evaluation of novel drug candidates in the IL-13tg mouse model which develops a human-like TB pathology. The infection of animals with M. tuberculosis, their treatment with antibiotics and the sampling of infected lung tissue are performed under biosafety level 3 (BSL-3) conditions. After inactivation of mycobacteria, mass spectrometry imaging (MSI) is used to measure the distribution of antibiotics in lung tissue. The concentration of drugs in defined pulmonary areas is determined by combining laser capture microdissection (LCM) and LC-MS/MS. The anti-mycobacterial activity within granulomas is analysed by LCM and subsequent molecular detection of mycobacteria (molecular bacterial load assay, MBLA).

Fig. 5. Validation of the novel antibiotic BTZ-043 in IL-13tg mice showing centrally necrotizing granulomas. (A) Spatial and temporal distribution of BTZ-043. The correlation of histopathology (HE staining, upper row) with the lesional distribution of BTZ-043 (MSI measurements, lower row) reveals an increased intensity of BTZ-043 in the rim of foamy macrophages of granulomas 2 hours after administration and a full penetration of centrally necrotizing granulomas at later time points (4 hours and 8 hours). (B) Quantification of BTZ-043 by using a combination of laser capture microdissection (LCM, top) and LC-MS/MS confirms the accumulation of BTZ-043 in necrotic granulomas, with concentrations several fold above the minimum inhibitory concentration. (C) Detection of the antimycobacterial activity of BTZ-043 in centrally necrotizing granulomas by combining LCM (top) and molecular M. tuberculosis detection (molecular bacterial load assay).

In a comprehensive research project, the Infection Immunology research group is investigating how the formation of necrotic granuloma lesions can be therapeutically influenced by targeted immunomodulation in IL-13tg and C3HeB/FeJ mice. By inhibiting key immunometabolic and inflammatory signalling pathways – including Arg-1, p38MAPK, NLRP3, ACC2, IL-27Rα and anaphylatoxin receptors – the inflammatory and metabolic environment within the granulomas is specifically altered, thereby attenuating necrotic processes and stabilising tissue architecture. This modulation not only leads to improved local immune balance but also facilitates the penetration of antibiotics into pathologically altered lesions, thereby increasing their effectiveness under disease-relevant conditions. While inhibition of these targets in C57BL/6 mice already leads to a significant reduction in pulmonary inflammation (Fig. 6), our IL-13tg and C3HeB/FeJ models will enable the investigation of these approaches in complex, human-like granulomas. The project thus combines the mechanistic elucidation of damaging immune responses with the development of adjuvant, immunomodulatory therapeutic strategies to improve existing antibiotic regimens.

Fig. 6. p38MAPK inhibition with doramapimod reduces tissue inflammation and reduces the bacterial load in chronically infected C57BL/6 mice. C57BL/6 mice were infected with 100 CFU M. tuberculosis. After 28 days, the mice were treated with a vehicle, doramapimod, isoniazid (INH) and rifampicin (RIF) or INH/RIF and doramapimod. After 56 days, (A) the number of granulomas and (B) the bacterial load in the lungs were quantified.