Spatial mapping reveals granuloma diversity and histopathological superstructure in human tuberculosis

Abstract

The hallmark of tuberculosis (TB) is the formation of immune cell-enriched aggregates called granulomas. While granulomas are pathologically diverse, their tissue-wide heterogeneity has not been spatially resolved at the single-cell level in human tissues. By spatially mapping individual immune cells in every lesion across entire tissue sections, we report that in addition to necrotizing granulomas, the human TB lung contains abundant non-necrotizing leukocyte aggregates surrounding areas of necrotizing tissue. These cellular lesions were more diverse in composition than necrotizing lesions and could be stratified into four general classes based on cellular composition and spatial distribution of B cells and macrophages. The cellular composition of non-necrotizing structures also correlates with their proximity to necrotizing lesions, indicating these are foci of distinct immune reactions adjacent to necrotizing granulomas. Together, we show that during TB, diseased lung tissue develops a histopathological superstructure comprising at least four different types of non-necrotizing cellular aggregates organized as satellites of necrotizing granulomas.

Figure S1.

Tissue landscape of the M.tb-infected lung. (A) H&E staining of lung tissue samples used in this study with matching multiplex IF staining with CD20, CD4, CD8, and CD68. Scale bars represent 5 mm. (B) Representative image of giant cells identified based on morphology and CD68 and DAPI staining. (C) Representative multiplex staining with CD20, CD4, and Ki-67 showing a Germinal center-like structure. (D) The same composite image of TB lung section as in Fig. 1 A stained with DAPI, with a representative view of segmented cells shown. (E) Cell density within tissue regions defined pathologically on H&E-stained tissue (n = 5). Dotted lines indicate density thresholds selected for tissue classification in Fig. 1, F and G.

Figure 1.

Active pulmonary TB is characterized by development of histopathological superstructures in the lung. (A) A TB lung tissue section stained with H&E with various histological and histopathological features indicated. (B) Enlarged example images of distinct lesion types identified in A. The outer and inner borders of necrotizing lesions were marked with solid and dotted red lines, respectively. The borders of non-necrotizing lesions were marked with solid black lines. (C) The proportion of non-necrotizing granulomas and lymphoid aggregates across 10 patients. Each color represents one type of non-necrotizing structure. (D) Composite Opal IF image of the same tissue section in A stained for major immune populations inlaid with magnified views of a necrotizing lesion (top panel, circled by a white line, *necrotizing core) and two non-necrotizing regions (bottom panel, rectangle boxes). (E) 3D diagram showing the variations in the density of DAPI⁺ segmented cells across entire TB lung section from A. The height of the peak (z axis) indicates cell density. (F) A representative 2D histogram showing a cross-section (indicated by the white dotted line in C) along y and z axis. The three z-axis segments (divided by dotted lines) are defined based on cell density and correspond to the regions of healthy, consolidated tissue and cellular aggregates shown in the pseudo-color topography image (right) of the same TB lung tissue as in A and C. Dotted lines indicate boundaries of necrotic areas. (G) Percentage of major immune cell populations among total DAPI-positive cells in CT (green) versus CA region (red). Each paired point represents a patient. The level of statistical significance is determined using a Wilcoxon signed-rank test and a P value <0.05 was considered statistically significant. *P < 0.05; **P < 0.01. (H) Distribution of immune cell-containing cell aggregates (colored) across the same tissue section as in A. The lesions selected are high in cellularity and positive for at least one of four major immune cell lineage markers (CD4, CD20, CD8, or CD68). Color intensity of non-greyed areas correlates inversely with immune cell density of the region relative to the maximum density in the tissue.

Figure 2.

Lesions have diverse compositions but fall across a continuous spectrum underpinned by CD20⁺ cell density. (A) Image of segmented immune cells in positively identified lesions in TB lung from Fig. 1 A with two representative areas (boxed) showing distinct cellular composition. (B) The density of each major immune cell population in necrotizing and non-necrotizing lesions from 13 patients. Each symbol represents an individual lesion and lines indicate group median. The level of statistical significance is determined using a Mann–Whitney test and a P value <0.05 was considered statistically significant. ***P < 0.001; ****P < 0.0001. (C) PCA plot of all lesions (n = 726) with necrotizing lesions indicated in red and circled by an ellipse. Additional marker-specific PCA plots are colored according to the density of marker-positive cells. (D) PCA plot visualization of all lesions across 10 patients. Each color indicates one patient, and each symbol represents an individual lesion. Images of example lesions of four individual patients are shown.

Figure 3.

Combined analysis of compositional and spatial features identifies four types of TB lesions. (A) Top: Images of DAPI-stained tissue showing variations in intra-lesion cell distribution. Bottom: Heatmap images showing the same three lesions with relative cell density indicated. (B) Left: A lesion image with the distance of each individual cell from the lesion border indicated by color density. Right: The same lesion with areas colored according to which cells are closer or further from the lesion border relative to the median distance. The equation used to calculate total cell central preference index is shown. (C) Schematic diagram illustrating the intra-lesion cell distribution patterns and their corresponding tCPI. (D) tCPI compared between necrotizing and non-necrotizing lesions. Each symbol represents an individual lesion. Bars indicate the group median. The level of statistical significance is determined using a Mann–Whitney test, ****P < 0.001. (E) Pairwise comparison of PC1 from Fig. 2 C against tCPI. Each symbol represents an individual lesion. Lesions (n = 726) are divided into four subpopulations with the quadrants being set at PC1 = 0 and tCPI = 0.5. The fifth type of lesions, necrotizing lesions (black symbols), are additionally indicated in the boxed area. (F) Proportion of each lesion type across 10 patients. Each color represents one lesion type. (G) Representative multiplex images of each lesion type with correlated images of DAPI and CD20 staining are shown below each multiplex image.

Figure S2.

Lesion cellular features in the lungs of TB patients. (A) Comparison of PC1 against tCPI from Fig. 3 E colored according to patient. (B) Comparison of PC1 against tCPI from Fig. 3 E colored according to the density of CD20⁺ B cells. (C) Comparison of CD20⁺ B cell density against tCPI. Each symbol represents an individual lesion, and lesions are colored according to type as Fig. 3 E. A value of 1 was added to the density of all lesions to allow lesions with no CD20⁺ B cells to be displayed on the log scale. (D) Density of major immune populations according to lesion type. Solid lines indicate the group median and dotted lines indicate interquartile range. (E) IF images stained for M.tb and DAPI with magnified views of the indicated boxed areas. Middle image represents the same tissue section as Fig. 4 E. (F) Distance of M.tb clusters from the necrotic interface in necrotizing lesions. M.tb was not identified in non-necrotizing regions. Each symbol represents a well-formed M.tb cluster (total five identified in the three lesions shown in A). (G) Proportion of major cell populations positive for Ki-67. Solid lines indicate the group median and dotted lines indicate interquartile range. (H) Necrotizing cell index of major immune populations. Each symbol represents an individual necrotizing lesion and lines track data from the same lesion. Solid red lines represent the group median. (I) Density of CD8⁺ and CD68⁺ cells in lesions located greater or less than 500 µm from necrotizing lesions. Solid lines indicate the group median and dotted lines indicate interquartile range. The level of statistical significance in I is determined using a Mann–Whitney test, and a P value <0.05 was considered statistically significant. **P < 0.01; ***P < 0.001. The level of statistical significance in D, G, and H is determined using a Kruskal–Wallis test with Dunn’s multiple comparisons test, and an adjusted P value <0.05 was considered statistically significant. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 4.

Spatial organization of CD68⁺ macrophages is lesion-type dependent. (A) A lesion image with CD68 staining (left) and a diagram illustrating the measurement of the closest distance of individual CD68⁺ cells from the lesion border (right). (B) The formula for calculating the spatial distribution of a cell-of-interest within a lesion. Schematic diagram of the lesions with representative immCPI measurements. (C) immCPI of each immune cell type across lesion type. Each symbol represents an individual lesion. Solid lines indicate group median and dashed lines indicate an immCPI of 1. (D) Image of a H&E-stained TB lung sample. (E) Pseudo-color image of the section from D with the necrotizing border indicated in blue. The magnified images of the boxed areas show positive M.tb staining and spatial position of CD68⁺, CD20⁺, and CD8⁺ cells. (F) Distance of individual immune cell population from necrotizing border in a representative lesion. Each symbol represents an individual cell and lines indicate group median. (G) Image of a H&E-stained TB lung sample. (H) Spatial distribution of monocyte/macrophage populations defined by CD14 and CD68 in tissue section from G, with the magnified image showing individual segmented myeloid cells. (I) Proportion of each macrophage subset by lesion type. (J) Distance of each macrophage subset from the closest necrotizing area across the entire tissue section. Each symbol represents an individual patient and column height indicates the group median. The level of statistical significance in C, F, and J is determined using a Kruskal–Wallis test with Dunn’s multiple comparisons test and an adjusted P value <0.05 was considered statistically significant. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 5.

Formation of non-necrotizing lesions is shaped by the neighboring necrotizing granuloma and intra-lesion organization of CD20⁺ and CD68⁺ cells. (A) Image of a H&E-stained TB lung sample. (B) The same TB lung section as in A overlaid with the locations of non-necrotizing and necrotizing lesions identified using LANDSCAPE tool. The relationship of each non-necrotizing lesion to the closest necrotizing lesion is indicated with connecting lines. (C) Distance of four types of non-necrotizing lesions to the closest necrotizing granuloma. Each symbol represents an individual non-necrotizing lesion and solid lines indicate the group median. The level of statistical significance is determined using a Kruskal–Wallis test with Dunn’s multiple comparisons test, and an adjusted P value <0.05 was considered statistically significant. **P < 0.01; ***P < 0.001; ****P < 0.0001. (D) The density of CD4⁺ and CD20⁺ cells in lesions located greater or less than 500 µm from necrotizing lesions. Solid lines indicate the group median and dotted lines indicate interquartile range. The level of statistical significance is determined using a Mann–Whitney test, and a P value <0.05 was considered statistically significant. **P < 0.01; ***P < 0.001. (E) Heatmap of inter-lesion trajectory analysis based on spatial and compositional parameters of individual lesions defined by LANDSCAPE. Lesions are ranked along the x axis according to trajectory and manually ordered on the y axis. Top panel indicates the lesion type. The analysis was performed using R. (F) Trajectory analysis of all lesions based on spatial and compositional lesion parameters as shown in E. The line indicates inferred trajectory. Each symbol represents an individual lesion (n = 726). (G) Schematic diagram summarizing key findings and illustrating hypothetical lesion progression pathways. Dotted lines indicate hypothetical lesion progression branchpoints and pathways. Question marks represent either lesion types or progression branchpoints yet to be formally determined. Schematic lesions identified in this study are depicted with color whereas those that have yet to be formally determined are shown in grey.