Inflammatory Macrophages Associate With Tissue Injury and Fibrosis in a Mouse Model of Tuberculosis

Abstract

Background: Post-tuberculosis lung disease causes a significant burden of global disease. While a consensus definition of post-tuberculosis lung disease is still in development, parenchymal cavitation, bronchiectasis, and fibrosis are recognized pathologic features. The molecular mechanisms driving development of each feature are largely unknown.

Methods: To facilitate the mechanistic study of tuberculosis-associated pathologic tissue remodeling and fibrosis, we adapted a mouse model of infection.

Results: The morphologies of fibrosis observed in mice were similar to those observed in human tissue samples, and fibrillar collagen deposition did not resolve with antituberculosis antibiotics. Inflammatory transcriptional signatures were persistently upregulated during chronic infection and did not fully resolve after weeks of antibiotics. Inflammatory and fibrosis-associated macrophages similarly persisted during treatment. Immunofluorescence microscopy revealed persistent macrophage populations and shifts in abundance and distribution of type 2 alveolar cells at sites of fibrogenesis.

Conclusions: A mouse model recapitulates key aspects of tuberculosis-assocaiated fibrosis. Transcriptional and cellular markers of inflammation persist through weeks of antibiotic treatment.

Keywords: fibrogenesis; fibrosis; host-pathogen; macrophage; mycobacteria; tissue injury; tuberculosis.

Figure 1.

Morphologies of tuberculosis-associated fibrosis in human disease. Samples from a human lung tissue repository were stained with hematoxylin and eosin (A top) or Masson’s trichrome (A bottom and B–C). A, Cavity wall with multiple non-necrotic, fibrotic, and fibro-calcific granulomas in the subpleural and parenchymal regions. B, Top, fibroinflammatory changes, including early alveolar replacement fibrosis with intra-alveolar chronic inflammatory cells and macrophages. Bottom, (arrowhead) early interstitial fibrosis and (arrow) intermediate pleural fibrosis. C, Top, intermediate interstitial fibrosis (arrowhead) and late pleural fibrosis (arrow). Bottom, late-stage parenchymal scarring/fibrosis effacing normal lung architecture.

Figure 2.

Mycobacterium tuberculosis-infected mice develop fibrotic histopathology similar to human disease. C3HeB/FeJ mice were infected with M. tuberculosis Erdman via the aerosol route. A, Mice were euthanized at the indicated time points postinfection and lungs were plated for colony-forming unit (CFU). B, Lungs were harvested and formalin-fixed at the indicated time points postinfection. Lungs were stained with hematoxylin and eosin to visualize inflammatory infiltrates; infiltrates were quantified using QuPath [27, 28]. C, Lungs were harvested at the indicated time points postinfection and stained with Masson’s trichrome to visualize collagen. Arrowheads, early interstitial fibrosis. Black bars, 100 μm; yellow bar, 200 μm. A, Error bars represent mean ± SD. B, Each point represents 1 mouse; bar represents the mean.

Figure 3.

Tuberculosis-associated fibrosis in a mouse model is quantifiable and does not resolve with antibiotics. C3HeB/FeJ mice were infected with Mycobacterium tuberculosis Erdman via the aerosol route. Infection was allowed to progress for 12 weeks; mice were then treated 6 times per week with RIPE therapy or water carrier control. A, Infection, treatment, and harvest schema. B, Lungs were harvested at the indicated time points and plated to quantify bacterial burden. C, Lungs were fixed at the indicated time points and stained with hematoxylin and eosin to visualize inflammatory infiltrates. QuPath was used to quantify inflammatory infiltrates relative to total lung area [27, 28]. One-way ANOVA followed by Tukey multiple comparison test. D, Second harmonic generation microscopy quantification of fibrillar collagen in carrier- and RIPE-treated mice over time. Each point represents 1 mouse; bars represent mean ± SD. Mann-Whitney U test. *P < .01, ***P < .001. Abbreviations: ns, not significant; RIPE, rifampin, isoniazid, pyrazinamide, and ethambutol.

Figure 4.

Transcriptional signatures of matrix remodeling and inflammation persist through antibiotic treatment. Mice were infected with Mycobacterium tuberculosis Erdman via the aerosol route with approximately 100 colony-forming units/mouse and treated as shown in Figure 3A. Lungs were harvested at the indicated time points; RNA was isolated from whole lung and comprehensive transcriptomic profiling was performed. A, Heatmaps of matrix remodeling, interferon-stimulated, and NF-κB-dependent genes changed ≥ 2-fold between uninfected and infected conditions. B–C, Reactome pathway analysis was used to identify the pathways differentially regulated between the indicated conditions. Abbreviation: IFN, interferon; IL, interleukin; Inf, infected; NF-κB, nuclear factor-κB; RIPE, rifampin, isoniazid, pyrazinamide, and ethambutol.

Figure 5.

Inflammatory macrophages and macrophage-associated fibrotic transcriptional signatures persist during antibiotic treatment. Mice were infected with Mycobacterium tuberculosis Erdman via the aerosol route with approximately 100 colony-forming units/mouse. A and D, Mice were treated and harvested at the indicated time points as shown in Figure 3A. A, Single-cell suspensions were generated and lungs were antibody stained to identify immune cell subsets [39]. Gating was performed as previously published [21]. UMAP representation of the identified populations. Band C, Infection was allowed to progress to 27 weeks postinfection. Single-cell suspensions were generated and lungs were antibody stained to identify immune cell subsets (gating strategy as in Supplementary Figure 4). D, Whole-lung RNA was used for comprehensive transcriptional profiling; expression of genes previously identified as reflecting a profibrotic macrophage population was profiled [38]. B, * P< .01, unpaired t test. Abbreviations: AM, alveolar macrophage; IM, interstitial macrophage; Inf, infected; macs, macrophages; MFI, mean fluorescence intensity; Mo-AM, monocyte-derived AM; NK cell, natural killer cell; RIPE, rifampin, isoniazid, pyrazinamide, and ethambutol; TR-AM, tissue-resident AM.

Figure 6.

Macrophages persist at sites of tissue remodeling during treatment. Mice were infected with Mycobacterium tuberculosis Erdman via the aerosol route with approximately 100 colony-forming units/mouse and treated as shown in Figure 3A. Lungs were harvested at the indicated time points. A–C, Top row, lungs were stained with Masson’s trichrome to visualize collagen. Bottom row, lungs were probed with antibody against F4/80 to visualize macrophages and DAPI to visualize all nuclei. D, Top row, lungs were stained with Masson’s trichrome to visualize collagen. Bottom row, immunohistochemistry staining for M. tuberculosis purified protein derivative (PPD) was used to visualize bacteria and bacterial products. Arrows: foamy macrophages staining positive for PPD. Scale bars as indicated. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; RIPE, rifampin, isoniazid, pyrazinamide, and ethambutol.

Figure 7.

Type 2 alveolar epithelial population shifts are spatially associated with fibrogenesis. Mice were infected with Mycobacterium tuberculosis Erdman via the aerosol route with approximately 100 colony-forming units/mouse and treated as shown in Figure 3A. Lungs were harvested at the indicated time points. A–C, Top row, lungs were stained with Masson’s trichrome to visualize collagen. Bottom row, lungs were probed with antibody to prosurfactant protein C to visualize AT2 cells. D, Resected human lung sample from an individual with active treated pulmonary tuberculosis was stained with Masson’s trichrome (left) and by immunohistochemistry for thyroid transcription factor-1 (right). Scale bars as indicated. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; RIPE, rifampin, isoniazid, pyrazinamide, and ethambutol.