Splenic ILCs are inhomogenously distributed. (A) Representative immunofluorescence image of WT splenic ILCs. Spleen sections were stained with the indicated markers, allowing identification of the major anatomical regions and ILCs subsets. NK cells were identified as CD3⁻NK1.1⁺Eomes⁺ cells (purple arrows), ILC1s were identified as CD3⁻NK1.1⁺Eomes⁻ cells (cyan arrows), ILC2s were identified as CD3⁻Gata3⁺ cells (green arrows), and ILC3s were CD3⁻Rorgt⁺ (orange arrows). n = 3 spleens. Scale bars, 200 µm for low-magnification images, 100 µm for zoomed-in insets. (B) Spatial frequency distribution of WT cDCs around NK1.1⁺Eomes⁺ ILC1s and Gata3⁺ ILC2s for the example shown in Fig. 3 A and median nearest neighbor distance. n = 3 spleens. Repeated-measures ANOVA with Tukey’s multiple comparisons post-test. (C) Median nearest neighbor distance of WT cDCs toward Rorgt⁺ ILC3s per spleen analyzed. Observed values for Sirpα⁻ cDC1s, Sirpα⁺CD4⁻ cDC2s, and Sirpα⁺CD4⁺ cDC2s are shown in gray, green, and purple, respectively. Distances between Rorgt⁺ ILC3s and randomly sampled Sirpα⁺CD4⁺ cDC2s are shown in cyan (n = 39); the boundaries of the simulated datasets are depicted by the dotted lines. (D) Median nearest neighbor distance of WT cDCs toward Rorgt⁺ ILC3s per spleen analyzed. Observed values for Sirpα⁻cDC1s, Sirpα⁺CD4⁻ cDC2s, and Sirpα⁺CD4⁺ cDC2s are shown in gray, green, and purple, respectively. Distances between randomly relabeled Rorgt⁺ ILC3s and Sirpα⁺CD4⁺ cDC2s are shown in cyan (n = 39); the boundaries of the simulated datasets are depicted by the dotted lines. (E) Probability of observing a cDC of a given subset as a function of spatial distance from Rorgt⁺ ILC3s in the two additional spleens analyzed. The observed probability is shown in bold. The result of randomly relabeling Rorgt⁺ ILC3s across the ILC repertoire is shown in gray (n = 39). The upper and lower boundary values for those simulations are depicted by the dotted lines.