Fig 1.
Schematic diagram of the two main organizational types of granuloma-associated fibrosis in TB granulomas indicating the two main types of fibrosis observed in NHP granulomas.
Yellow cells represent myofibroblasts, green represent macrophages, and grey fibroblasts.
Fig 2.
Macrophage polarization and differentiation pathways in GranSim.
A: Representation of signaling pathways for regulatory macrophages and classification of signals as M1 (inflammatory) or M2 (anti-inflammatory). Every time a macrophage receives a signal on the specific pathway a counter is incremented representing the total stimulation level (e.g. stat1Counter). B: Schematic diagram of criteria for macrophage to myofibroblast transformation (MMT) that were implemented in GranSim following, namely that resting macrophages that have STAT1 and STAT3 signals and are not exposed to Mtb.
Table 1.
Parameters used in GLM for predicting the location of a macrophage.
Locations were coded as either peripheral (1) or central (0) depending on their proximity to the granuloma boundary (see Methods for more details).
Table 2.
Parameter names, definitions and ranges used for LHS.
Fig 3.
Macrophage locations within simulated granulomas at 50 days.
Cells were characterized as “central” if they were located less than 80% of the radius of the granuloma away from center or “peripheral” if they were between 80% and 100% of the radius. Remaining cells were categorized as “Not in granuloma”.
Fig 4.
Granulomas have spatially-distinct patterns of STAT1 and STAT3 phosphorylation, including phospho-STAT1+STAT3+ cells.
A macaque granuloma was stained for p-STAT1 (green) and p-STAT3 (red) and CD163 (blue). ImageJ was used to separate the (A) merged channels into (B) p-STAT3 and (C) and p-STAT1 channels. (D) ImageJ’s image calculator function was used with the AND command to identify pixels where p-STAT1 and p-STAT3 are colocalized (yellow). CD163 (blue) was included in each panel for context. Scale bar represents 200 μm.
Fig 5.
Four distinct types of fibrosis phenotypes produced by the hypothesized MMT rules.
Plots show the macrophages that meet the STAT1+STAT3+Exposure- criteria for MMT (turquoise), or not (coral), at 50 days post-infection. The four fibrosis types are A: central fibrosis, B: peripheral fibrosis, C: fully fibrotic, and D: non-fibrotic. E: Average time between STAT1 phosphorylation signals in GranSim simulations. STAT1 signal comes before STAT3 signal in macrophages that are candidates for MMT.
Fig 6.
Snapshots of GranSim simulations show fibrosis occurring with MMT present.
Four different granulomas and fibrosis phenotypes are shown at days 50 and 100 post-infection with Mtb. A,B: No fibrosis, C,D: peripheral fibrosis, E,F: central fibrosis, G,H: fully fibrotic. Cells are represented by different colors as follows: resting macrophages are green, activated macrophages are blue, infected macrophages are orange, chronically infected macrophages are red, IFNγ producing T-cells are pink, cytotoxic T cells are violet, regulatory T cells are cyan, fibroblasts are maroon, and myofibroblasts are gold. I: Fibrosis significantly lowers the total cell count in a virtual granuloma. Total cell counts per granuloma snapshot 100 days after the initial Mtb infection are shown on a log10 scale.
Fig 7.
Characterization of fibrosis in the granuloma.
A-D: IHC stains for α SMA (left panels) and corresponding Geographically Information System (GIS) analysis-identified locations of fibroblasts (right panels) for four unique non-human primate granulomas. Objects identified as fibroblasts are highlighted in cyan, and the granuloma boundary demonstrated by the solid yellow line. E, F: Distribution of lengths (E) and widths (F) of αSMA+ cells for each granuloma respectively. G: Total numbers of αSMA+ cells for each granuloma, respectively. Cells with length or width of greater than three standard deviations away from the mean size were classified as outliers and removed from the plots in E-G.H-J: Representative IHC staining of fibrotic granuloma demonstrate that αSMA (green) expressing cells in the fibrotic cuff co-express CD11c (red), vimentin (blue), and CD31 (cyan). H: Expected expression levels of markers shown in panel I for each cell type involved in fibrosis. Colors correspond to the stains used in B and semi-opaque represents low expression. I: Full staining on sample C. J: Inset, αSMA, vimentin, CD11c, CD31 (left to right).
Fig 8.
Snapshots of simulation show fibrosis occurring with both MMT and lung-resident fibroblast involvement.
A,B: None fibrotic, C,D: peripheral fibrosis, E,F: central fibrosis, G,H: fully fibrotic. Cells are represented by different colors as follows: resting macrophages are green, activated macrophages are blue, infected macrophages are orange, chronically infected macrophages are red, IFNγ producing T cells are pink, cytotoxic T cells are violet, regulatory T cells are cyan, fibroblasts are maroon, and myofibroblasts are gold.
Fig 9.
Fibroblast numbers in simulated 2D granulomas (log10) comparing where (A) both fibroblasts to myofibroblast differentiation together with MMT or (B) MMT alone.
Number of myofibroblasts and activated fibroblasts are shown under A: fibroblasts and MMT involvement in fibrosis, B: MMT alone in fibrosis. Plots are shown on a log base 10 scale. C: Sensitivity analysis reveals key important parameters for increasing the number of myofibroblasts. PRCCs were calculated as described in Methods section and range between -1 and 1. Shown are only the significant PRCC values when the outcome variable is number of myofibroblasts for p < 0.05. Parameters are defined and explained in Table 2.