Microanatomic Distribution of Myeloid Heme Oxygenase-1 Protects against Free Radical-Mediated Immunopathology in Human Tuberculosis

Abstract

Heme oxygenase-1 (HO-1) is a cytoprotective enzyme that controls inflammatory responses and redox homeostasis; however, its role during pulmonary tuberculosis (TB) remains unclear. Using freshly resected human TB lung tissue, we examined the role of HO-1 within the cellular and pathological spectrum of TB. Flow cytometry and histopathological analysis of human TB lung tissues showed that HO-1 is expressed primarily in myeloid cells and that HO-1 levels in these cells were directly proportional to cytoprotection. HO-1 mitigates TB pathophysiology by diminishing myeloid cell-mediated oxidative damage caused by reactive oxygen and/or nitrogen intermediates, which control granulocytic karyorrhexis to generate a zonal HO-1 response. Using whole-body or myeloid-specific HO-1-deficient mice, we demonstrate that HO-1 is required to control myeloid cell infiltration and inflammation to protect against TB progression. Overall, this study reveals that zonation of HO-1 in myeloid cells modulates free-radical-mediated stress, which regulates human TB immunopathology.

Keywords: free radical; heme oxygenase-1; histopathological spectrum; human TB pathology; human pulmonary tuberculosis; karyorrhexis; macrophage; mycobacterium tuberculosis; myeloid cell inflammation; neutrophil.

Graphical abstract

Figure 1

HO-1 Staining Profile in the Cavity Wall HO-1 staining of cellular component in adluminal cells with bright staining of phagocytic giant cells (arrowheads, inset i), bright staining of neutrophils (arrows, inset ii), and negative karyorrhectic neutrophil staining (rectangle, inset iii) (L = Lumen). Shown is bright staining of histiocytes in the granulomatous layer (black asterisks) and bright staining of inflammatory cells (lymphocytes, plasma cells, and histiocytes) and endothelial cells (yellow arrows) in the granulation tissue layer (asterisk).

Figure 2

HO-1 Staining of Tubercles and Non-necrotizing Granulomas (A and B) Low-power magnification of HO-1 staining in lung parenchyma with multiple tubercles (T) (A) and high-power demonstration of HO-1 positivity in scattered neutrophils (arrows) in the central caseative component (B). Also shown are HO-1-negative granular debris (square, inset), HO-1-positive giant cells, and epithelioid histiocytes in granulomas (arrowheads) and HO-1-positive endothelial cells lining capillaries in fibrous lamellae (oval). (C and D) Bright HO-1 staining of granulomas (rectangle) at medium (C) and high magnification (D). Langhans giant cells (D, arrows) and epithelioid histiocytes (D, black arrowheads) were strongly HO-1 positive. HO-1-positive endothelial cells (yellow arrowheads) in the adjacent vasculature and myofibroblasts (oval) are shown.

Figure 3

HO-1, ROS, and RNS Levels in Myeloid Cells Isolated from Severely Diseased Regions of Human Tuberculous Lung Tissue (A) The sequential flow cytometry gating strategy for identification of monocytes, macrophages, neutrophils, and T cells from freshly isolated CD45⁺ cells within the distinct pathological regions of human TB lung. Cells were identified by the following cell-surface markers: monocytes (CD14⁺HLA-DR⁺CD11c^lowCD66b⁻CD3⁻), macrophages (HLA-DR⁺CD11c⁺CD206⁺CD11b⁺CD86⁺CD16⁻CD66b⁻CD3⁻), neutrophils (CD11b⁺CD66b⁺CD16⁺CD14⁻CD3⁻), and T cells (CD3⁺CD14 ⁻CD11b⁻). (B) Representative histogram of HO-1 expression in neutrophils. (C and D) Percentage of HO-1⁺ neutrophils (C) and MFI of HO-1 in neutrophils (D). (E) Representative histogram of HO-1 expression in macrophages. (F and G) Percentage of HO-1⁺ macrophages (F) and MFI of HO-1 in macrophages (G). (H) Representative histogram of HO-1 expression in monocytes. (I and J) Percentage of HO-1⁺ monocytes (I) and MFI of HO-1 in monocytes (J). (K and L) Percentage of ROS-positive neutrophils, macrophages, and monocytes (K) and their respective MFI of ROS (L). (M and N) Percentage of RNS-positive neutrophils, macrophages, and monocytes (M) and their respective MFI within isolated lung immune cells (N). n = 9–12 individual patients per group (each dot represents an individual patient). Statistical testing was performed using the unpaired Student’s t test. Data are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001. Black curve in (B), (E), and (H) represents isotype control (Iso).

Figure 4

HO-1^−/− Mice Are More Susceptible to Mtb Infection (A) Kaplan-Meier survival analysis of HO-1^+/+ and HO-1^−/− mice following infection with Mtb H37Rv. Uninfected HO-1^+/+ and HO-1^−/− mice were used as controls and survived the duration of the study (data not shown). n = 10 mice per group. (B and C) Mtb bacillary burden in lungs (B) and spleens (C) of HO-1^+/+ and HO-1^−/− mice at 6, 12, and 18 weeks post-infection. (D) H&E staining of representative lung and spleen sections from Mtb-infected HO-1^+/+ and HO-1^−/− mice at 18 weeks post-infection. (E) Trichrome staining of representative lung and spleen sections from Mtb-infected HO-1^+/+ and HO-1^−/− mice at 18 weeks post-infection. Blue (as indicated by arrows) shows collagen deposition (scale bar, 100 μm). (F) Western blot analysis of HO-1 and Nrf-2 in the total lung protein extract of uninfected and Mtb-infected HO-1^+/+ and HO-1^−/− mice at 6, 12, 18, and 40 weeks post-infection. Each lane represents individual mice. (G and H) Quantification of HO-1 (G) and Nrf-2 (H) western blots shown in (F) using ImageJ software. Target genes were normalized against the housekeeping gene GAPDH. (I) Relative HO-1 mRNA expression in the lungs of Mtb-infected HO-1^+/+ mice was determined using qRT-PCR, normalized to the housekeeping gene β-actin. n = 4 for each time point. Statistical testing was performed using the unpaired Student’s t test. Data are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001.

Figure 5

Dysregulated Immune and Inflammatory Responses in Mtb-Infected HO-1^−/− Mice (A–M) Cytokine analysis of IL-1β (A), IL-2 (B), IL-3 (C), IL-5 (D), IL-6 (E), IL-10 (F), IL-17A (G), MIP-1α (H), MIP-1β (I), KC (J), GM-CSF (K), G-CSF (L), TGF-β (M) in the BALF of uninfected and Mtb-infected HO-1^+/+ and HO-1^−/− mice at 6, 12, and 18 weeks. (N and O) Percentage of neutrophils in the BALF (N) and lungs (O) of Mtb-infected HO-1^+/+ and HO-1^−/− mice at 6, 12, and 18 weeks post-infection. (P–S) Other monocytes and granulocytes in the lungs of Mtb-infected HO-1^+/+ and HO-1^−/− mice. (P and Q) Representative scatterplot for CD11b⁺Ly6C⁺F4/80⁺ monocytes (P) and percent differences (Q) at 6 weeks post-infection. (R and S) Representative scatterplot for CD11b⁺Ly6G⁺F4/80⁺Gr-1⁺ granulocytes (R) and percent differences at 12 and 18 weeks post-infection (S). n = 4 for each time point. Statistical testing was performed using the unpaired Student’s t test. Data are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001.

Figure 6

Myeloid Specific HO-1 Knockout Mice (HO-1^LysM−/−) Are Susceptible to Mtb Infection (A) Kaplan-Meier survival analysis of Mtb-infected HO-1^LysM+/+ and HO-1^LysM (n = 10 mice per group). (B and C) Bacterial burden in lungs (B) and spleens (C) of Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 6, 12, and 24 weeks post-infection. (D) Flow cytometry gating strategy for identification of myeloid cells. (E and F) Absolute numbers in cells per gram of tissue (E) and proportion of CD45⁺ cells for lung neutrophils (F) from uninfected or Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 12 and 24 weeks post-infection. (G and H) Absolute numbers in cells per gram of tissue (G) and proportion of CD45⁺ cells for lung macrophages (H) from uninfected or Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 12 and 24 weeks post-infection. (I) The sequential flow cytometry gating strategy for identification of IFN-γ-producing lymphocytes and T cells. (J) Representative gated population of CD45⁺IFN-γ⁺ T cells. (K and L) Differences in proportions of CD45⁺ cells for IFN-γ⁺ T cells (K) and IFN-γ MFI (L) isolated from the lungs of Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 24 weeks post-infection. (M) Representative gated population of CD45⁺CD4⁺IFN-γ⁺ T cells. (N) Differences in proportions of CD45⁺ cells for CD4⁺ T cells in the lungs of Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 24 weeks post-infection. (O and P) Differences in proportions of CD45⁺ cells (O) and IFN-γ MFI (P) for CD45⁺CD4⁺IFN-γ⁺ T cells isolated from the lungs of Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 24 weeks post-infection. (Q) Representative gated population of CD45⁺CD4⁺CD69⁺IFN-γ⁺ T cells. (R) Differences in proportions of CD45⁺ cells for CD45⁺CD4⁺CD69⁺ T cells in the lungs of Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 24 weeks post-infection. (S and T) Differences in proportions of CD45⁺ cells (S) and IFN-γ MFI (T) for CD45⁺CD4⁺CD69⁺IFN-γ⁺ T cells isolated from the lungs of Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 24 weeks post-infection. n = 4 for each time point. Statistical testing was performed using the unpaired Student’s t test. Data are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001.

Figure 7

Cytokine Response in HO-1 LysM Mice and Proposed Model for HO-1-Mediated Protection against Mtb Infection (A–K) Cytokine analysis (A), IL-2 (B), IL-3 (C), IL-5 (D), IL-6 (E), IL-10 (F), IL-17A (G), MIP-1α (H), MIP-1β (I), G-CSF (J), KC (K) in the BALF of uninfected and Mtb-infected HO-1^LysM+/+ and HO-1^LysM−/− mice at 6, 12, and 18 weeks post-infection. n = 4 for each time point. Statistical testing was performed using the unpaired Student’s t test. Data are presented as mean ± SEM. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.001. (L) Proposed model for HO-1-mediated protection during Mtb infection. Mtb infection induces HO-1 expression and its downstream enzymatic cascade. Together with its enzymatic products and ferritin, which is induced by HO-1, HO-1 protects against TB immunopathology and thereby contributes to overall TB disease tolerance via its anti-oxidant, anti-inflammatory, and anti-proliferative properties. Therefore, maintaining the physiological levels of HO-1 is important to limit TB disease pathology (green shaded area). Contrarily, in HO-1-deficient conditions, the levels of key pro-inflammatory cytokines are significantly elevated. This results in rapid and dysregulated myeloid cell infiltration to the infected sites, which results in significantly elevated levels of ROS and RNS as well as rapid karyorrhexis and NETosis. ROS and RNS also generate increased localized concentrations of peroxynitrite (ONOO⁻) and subsequent suppression of T cell responses. In addition, Mtb infection also causes endothelial injury and intravascular coagulation, resulting in significant accumulation of heme, a potent proinflammatory and pro-oxidant molecule. Excess heme overwhelms the cytoprotective activity of HO-1, thereby contributing pro-oxidant-mediated inflammatory responses. Together, these dysregulated inflammatory responses result in chronic TB immunopathology and rapid disease progression (red shaded area).