Claudin 1 facilitates positioning and contact between DC1s and mTECs. (A) Flow cytometry gating strategy of mTEC subsets. TECs were gated as EPCAM⁺CD45⁻ and further distinguished to LY51⁺UEA⁻ cortical thymic epithelial cells (cTECs) and LY51⁻UEA⁺ mTECs. The latter were further separated into AIRE⁺ITGB4⁻ mTEC^HI, AIRE^–ITGB4⁺ mTEC^LO, and double-negative cells, which were comprised of keratinocyte mimetics (LY6D⁺) and other mimetic cells (LY6D⁻). Defa6⁺ mTECs are color-coded in orange denoting their TdTOMATO positivity. (B) Frequency of CLAUDIN 3⁺ cells within cell populations from Fig. S3 I (mean ± SEM, n = 9 mice from three independent experiments). (C) Frequency of CLAUDIN 3⁺ cells and MFI of CLAUDIN 3 expression within mTEC subsets from Fig. S3 J (mean ± SEM, n = 5–9 mice from three independent experiments). (D) Schematic of competitive BM chimera used to assess the role of Claudin 1 in positioning of DC1 lineage cells in the proximity of mTECs. Mouse models used as donors of BM are marked by the letters a and b. Note that Adig^GFP recipients obtained BM either from mice a (BM mix a) or b (BM mix b). (E) Flow cytometry gating strategy used to analyze the reconstitution of competitive BM chimeras from Fig. 4 D. Thymic CD11c⁺XCR1⁺ cells (DC1s) were gated as in Fig. S4 F and either as TdTOMATO⁺ or TdTOMATO⁻. Note that TdTOMATO⁺ DC1s are Cldn1^+/+ and TdTOMATO⁻ DC1s are Cldn1^−/− in the case of mice receiving BM mix a, and TdTOMATO⁺ DC1s are Cldn1^−/− and TdTOMATO⁻ DC1s are Cldn1^+/+ in the case of mice receiving BM mix b. (F) Light sheet fluorescence microscopy images of analogous regions within the thymic medulla of competitive BM chimeras from Fig. 4, D and E. The top images capture the entire medullary compartment imaged. The bottom images visualize segmented objects (red, DC1s; and green, mTEC clusters) within selected regions of the whole 3D images shown above. Separate legends are shown for BM mix a and b. (G) Schematic of the analysis of the regions imaged in Fig. 4 F. The imaged area of each mTEC cluster captured was expanded by 50, 25, or 5 μm, and the percentage of DC1s from the total within these expanded clusters was counted. Note that DC1s localized up to 5 μm from mTEC clusters are considered to be in direct contact with mTECs (left panel). The percentage of DC1s that are within a defined distance of mTEC clusters related to Fig. 4 F. The number above the Cldn1^−/− DC1 columns indicates the fold change reduction in the percentage of Cldn1^−/− DC1 with respect to the percentage of Cldn1^+/+ DC1 (right panel). (H) Violin plots showing minimum distance (μm) between DC1 and the nearest mTEC cluster related to Fig. 4 F. Medians and quartiles are shown (n = 2,259 Cldn1^+/+ DC1s and 531 Cldn1^−/− DC1s per representative experiment from a total of two experiments). Statistical analysis in B, C, and H was performed using unpaired, two-tailed Student’s t test, ***P ≤ 0.001, ****P < 0.0001, ns = not significant. All mice were bred on the B6 background. Littermates were used as controls.