Spatial separation of CD4 ⁺ T cells and neutrophils in TB lesions. C57BL/6 (B6) and C3HeB/FeJ (C3H) mice were aerosol infected with M. tuberculosis HN878 and lungs harvested for multiparameter immunofluorescence staining of formaldehyde-fixed, paraffin-embedded sections at the indicated time points. (a) Representative low-power images of lung lobes, showing distribution of lesions. White boxes indicate areas at days 14 and 21 shown at greater magnification in panel d. The images from days 14 and 21 are reproduced in Fig. S3 a, along with those from the other replicate mice at these time points. Scale bars = 1 mm. (b) Quantification of numbers of lesions in whole lungs, normalized for tissue area. (c) Median area of lesions detected per mouse. (d) Representative images of lesions at 14 and 21 days after infection, showing all immune cell markers. Images are shown at different scales to aid visualization (day 14 scale = 20 µm; day 21 scale = 50 µm). (e) Number of cells positive for the indicated markers within lesions, normalized for the total area of all lesions. (f) Numbers of CD4⁺ T cells in contact with a macrophage annotation (≤0 μm distance), normalized for the total area of all lesions. (g) Stacked bar plots showing percentages of lesions across whole lungs falling into low (≤20%), intermediate (Int, >20% ≤40%), or high (Hi, >40) bins for coverage with Ly6G staining. Data shown are means ± standard error of all mice with detectable lesions (N = 3 for day 14 C3HeB/FeJ; N = 4 for others). (h) Representative images showing the relative distribution of CD4⁺ T cells and Ly6G staining in lesions with low and high Ly6G coverage. Scale bar = 50 μm. (i) Number of CD4⁺ T cells within lesions in the different Ly6G coverage bins at 21 days after infection, normalized for the total area of lesions analyzed per bin. Plots in b, c, e, f, and i show points representing all individual replicate mice with detectable lesions, with lines at the mean. Data shown are from a single experiment with N = 4 mice per group and are representative of two independent experiments. Statistical analysis: b, c, and e, two-way ANOVA with Holm–Sidak post hoc test; f, unpaired t test; g, Dirichlet-multinomial regression, with the indicated P values corresponding to the mouse strain effect on frequency of Ly6G-high lesions; i, one-way ANOVA with Holm–Sidak post hoc test. Actual adjusted P values are shown or: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. See also: Fig. S3, a–c.