Aire-dependent production of XCL1 mediates medullary accumulation of thymic dendritic cells and contributes to regulatory T cell development

Abstract

Dendritic cells (DCs) in the thymus (tDCs) are predominantly accumulated in the medulla and contribute to the establishment of self-tolerance. However, how the medullary accumulation of tDCs is regulated and involved in self-tolerance is unclear. We show that the chemokine receptor XCR1 is expressed by tDCs, whereas medullary thymic epithelial cells (mTECs) express the ligand XCL1. XCL1-deficient mice are defective in the medullary accumulation of tDCs and the thymic generation of naturally occurring regulatory T cells (nT reg cells). Thymocytes from XCL1-deficient mice elicit dacryoadenitis in nude mice. mTEC expression of XCL1, tDC medullary accumulation, and nT reg cell generation are diminished in Aire-deficient mice. These results indicate that the XCL1-mediated medullary accumulation of tDCs contributes to nT reg cell development and is regulated by Aire.

Figure 1.

Screening for chemokines that regulate the localization of tDCs. (A) RT-PCR analysis of chemokine receptor expression in isolated tDCs. Shown are the results of ethidium bromide detection of electrophoretically separated PCR products. Hprt, hypoxanthine-guanine phosphoribosyltransferase. Shown are the representative data of three independent experiments. (B) Chemotactic indexes of tDCs to the ligands of expressed chemokine receptors CCL1 (CCR8 ligand), CCL3 (CCR5 ligand), CCL7 (CCR2 ligand), CCL8 (CCR5 ligand), CCL17 (CCR4 ligand), CCL19 (CCR7 ligand), CCL20 (CCR6 ligand), CCL21 (CCR7 ligand), CCL22 (CCR4 ligand), CXCL1 (CXCR1 ligand), CXCL9 (CXCR3 ligand), CXCL10 (CXCR3 ligand), CXCL11 (CXCR3 ligand), CXCL12 (CXCR4 ligand), XCL1 (XCR1 ligand), and CX3CL1 (CX3CR1 ligand) were determined. Bar graphs indicate means ± standard errors of the data from at least three independent measurements. Red underlines indicate the chemokines that attracted the tDCs (based on the statistical significance as shown in the graph). *, P < 0.05; **, P < 0.01; ***, P < 0.001. (C) Frozen thymus sections from CCR7 ligand–deficient (plt/plt) and CXCR4-deficient (Mx-Cre x CXCR4^f/f) mice, as well as the control mice, were stained with anti-CD11c antibody (green) and mTEC-specific antibody ER-TR5 (red). The majority, if not all, of Cxcr4 was deleted in the thymus cells of Mx-Cre x CXCR4^f/f mice. C, cortex; M, medulla. Shown are the representative results of three independent experiments.

Figure 2.

Expression of Xcr1 and Xcl1 in the thymus. (A–C) Quantitative RT-PCR analysis of Xcr1 expression in sorted thymus cell populations, including CD45⁺ thymocytes (CD45⁺), CD45⁻ thymic stromal cells (CD45⁻), CD45⁻I-A⁺UEA1⁺ mTECs, CD11b⁺CD11c⁻ macrophages (MΦ), and CD11c⁺ tDCs (A); in sorted tDC subpopulations, including CD11c⁺ total tDCs, I-A^highCD11c^highCD11b⁻ lymphoid tDCs (lyDC), I-A^highCD11c^highCD11b⁺ myeloid tDCs (mDC), and I-A^mediumCD11c^mediumB220⁺ plasmacytoid tDCs (pDC) (B); and in CD11c⁺ DCs from the thymus (Thy), the spleen (Spl), the subcutaneous lymph nodes (subLN), and the mesenteric lymph nodes (mesLN; C). The amounts of Xcr1 were normalized to the amount of Hprt, and those in CD45⁺ thymocytes (A) and total tDCs (B and C) were arbitrarily set to 1. (D and E) Quantitative RT-PCR analysis of Xcl1 expression in thymus cell populations, including CD45⁺ thymocytes (CD45⁺), CD11c⁺ DCs, CD45⁻I-A⁻MTS15⁺ fibroblasts (Fib), CD45⁻I-A⁺ total TECs (TEC), CD45⁻I-A⁺Ly51⁺ cTECs (cTEC), and CD45⁻I-A⁺UEA1⁺ mTECs (mTEC); and in CD45⁻I-A^lowUEA1⁺ (MHC II^lo) and CD45⁻I-A^highUEA1⁺ (MHC II^hi) mTEC subpopulations. The amounts of Xcl1 were normalized to the amount of Hprt, and those in CD45⁺ thymocytes (D) and CD45⁻ I-A^lowUEA1⁺ mTECs (E) were arbitrarily set to 1. Bar graphs show means and standard errors of at least three independent measurements. (F) Quantitative RT-PCR analysis of Xcl1 expression in CD45⁻I-A⁺UEA1⁺ mTECs and anti-CD3–stimulated CD8⁺ T cells. Means (bars) and individual measurements (circles) are shown (n = 3).

Figure 3.

Distribution of DCs and macrophages in the thymus of Xcl1-deficient mice. (A) Two-color immunofluorescence analysis of thymus sections stained for CD11c (green) and the mTEC-specific ER-TR5 determinant (red). Representative images from three independent experiments are shown. White lines indicate borders among the indicated regions of the thymus section identified as in Fig. S2 A. C, cortex; M, medulla. Representative images from three independent experiments are shown. (B) Number of CD11c⁺ cells per unit area (1 mm²) of the indicated regions of the thymus sections was measured. Means and standard errors of cell numbers from three independent measurements are shown. (C) Means and standard errors (n = 6) of the absolute numbers of lymphoid DCs (lyDCs), myeloid DCs (mDCs), and plasmacytoid DCs (pDCs) in the thymus of indicated mice are shown. (D) Relative density of CD11c⁺ cells in the indicated regions to the density in the medullary region of the thymus section (in percentile). Means and standard errors (n = 3) for WT, plt/plt, Mx-Cre × CXCR4^f/f, Xcl1-deficient, and Xcl1-deficient plt/plt mice are plotted. (E and F) The thymus sections were analyzed for CD11b (green) and the ER-TR5 determinant (red). Representative images from three independent experiments are shown in E, and means and standard errors of cell numbers from three independent analyses are shown in F. ***, P < 0.001. NS, not significant (P > 0.05).

Figure 4.

Negative selection of thymocytes in Xcl1-deficient mice. (A) Thymocytes isolated from WT and Xcl1-deficient (Xcl1-KO) mice of C57BL/6 and BALB/c backgrounds were three-color stained for CD4, CD8, and indicated TCR-Vβs. Monoclonal antibodies specific for Cβ, Vβ3, Vβ5, and Vβ11 used were H57-597, KJ25, MR9-4, and KT11, respectively. Shown are the means and standard errors of the frequencies of indicated Vβ^high cells within CD4⁺CD8⁻ and CD4⁻CD8⁺ thymocytes from three independent measurements. *, P < 0.05. (B) Thymocytes from indicated mouse strains were three-color stained for CD4, CD8, and 2C-TCR (clone 1B2). Shown are the representative profiles of 2C-TCR–expressing cells from three independent experiments. (C) T cell–depleted bone marrow cells from OT-I-TCR–transgenic mice (left) or OT-II-TCR–transgenic mice (right) were transferred into lethally irradiated WT, RIP-mOVA–transgenic (RIP-mOVA), or RIP-mOVA–transgenic Xcl1-deficient (RIP-mOVA Xcl1-KO) H-2^b mice. Indicated thymocyte populations (CD4⁺CD8⁺, DP; CD4⁺CD8⁻, CD4SP; and CD4⁻CD8⁺, CD8SP) were stained for TCR-Vα2 (solid lines). Shaded profiles represent the analysis with control antibodies. Numbers indicate the frequency of Vα2^high cells within indicated populations. Representative results of three independent experiments are shown.

Figure 5.

nT reg cell development in Xcl1-deficient mice. (A) Thymocytes from indicated mouse strains were four-color stained for CD4, CD8, CD25, and intracellular Foxp3. Numbers indicate the frequency of cells within indicated areas. Shown are the representative results of four independent experiments. (B) Means and standard errors (n = 3–11) of the absolute numbers of CD4⁺CD8⁻CD25⁺Foxp3⁺ thymocytes in indicated mice at indicated ages. (C) Immunofluorescence analysis of thymus sections from indicated mice for Foxp3. Letters and lines indicate the regions identified as in Fig. S2 A. (D) Numbers of Foxp3⁺ thymocytes per unit area (1 mm²) of the indicated regions. Means and standard errors of cell numbers from three independent measurements are shown. (E) Immunofluorescence analysis of thymus sections from indicated mice for Foxp3 (green), CD11c (red), and mTECs (ER-TR5; blue). Magnified images in the M region are shown. (F) Frequency (percentage) of CD11c⁺ cell–attached cells among Foxp3⁺ cells in indicated thymic regions. Means and standard errors of the frequencies from eight different images (bars) and the numbers of CD11c⁺ cell–attached Foxp3⁺ cells (top) and total Foxp3⁺ cells (bottom) counted are shown. (G) Representative CFSE fluorescence profiles of CD4⁺CD25⁻ lymph node T cells from B6-Ly5.1 mice cultured in the absence or presence of CD4⁺CD8⁻CD25⁻ or CD4⁺CD8⁻CD25⁺ thymocytes from WT B6 (top) or Xcl1-deficient B6 background mice (bottom) with or without anti-CD3 antibody. Numbers indicate the frequency of CFSE^low cells. Representative results of three independent experiments are shown. (H) Intracellular Helios and Foxp3 expression of CD4⁺ spleen cells from WT and Xcl1-KO mice was analyzed by flow cytometry. Shown are the means and standard errors (n = 3) of cell numbers of Foxp3⁺, Foxp3⁺Helios⁺, and Foxp3⁺Helios⁻ CD4⁺ spleen cells from WT and Xcl1-KO mice. *, P < 0.05; ***, P < 0.001. NS, not significant (P > 0.05).

Figure 6.

Thymocytes from Xcl1-deficient mice elicit inflammatory lesions in lacrimal glands in nude mice. 5 × 10⁷ thymocytes isolated from WT or Xcl1-deficient (Xcl1-KO) mice of BALB/c background were intravenously transferred into BALB/c-nu/nu mice. Where indicated, equal numbers of thymocytes (5 × 10⁷ + 5 × 10⁷) were mixed before the transfer. Paraffin-embedded sections of lacrimal glands at 8 wk after the transfer were stained with hematoxylin and eosin. (A) Representative images of the sections at two different magnifications. (B) Histological scores of inflammatory lesions in the lacrimal glands (n = 5). Horizontal bars indicate the means. ***, P < 0.001.

Figure 7.

DCs and T reg cells in the thymus of Aire-deficient mice. (A) Quantitative RT-PCR analysis of Xcl1, Ccl19, Ccl21, and Ccl25 expression in sorted CD45⁻EpCAM⁺UEA1⁺ mTECs from WT and Aire-deficient mice. The amounts of the transcripts were normalized to the amount of housekeeping Hprt, and those in WT mTECs were arbitrarily set to 1. Bar graphs show means and standard errors of at least three independent measurements. (B) Immunofluorescence analysis of thymus sections for CD11c (green) and mTECs (ER-TR5; red). Representative images from two independent experiments are shown. Lines indicate borders among the indicated regions identified as in Fig. S2 A. (C) Numbers of CD11c⁺ cells per unit area (1 mm²) of the indicated regions were measured. Means and standard errors of cell numbers are shown. (D) Means and standard errors (n = 3–4) of the absolute numbers of indicated DC subpopulations in the thymus of indicated mouse strains are shown. (E) Immunofluorescence analysis of thymus sections from indicated mice for Foxp3. Representative images from two independent experiments are shown. (F) Numbers of Foxp3⁺ cells per unit area (1 mm²) of the indicated areas were measured. Means and standard errors of cell numbers from three independent measurements are shown. (G) Means and standard errors (n = 3) of the absolute numbers of CD4⁺CD8⁻CD25⁺Foxp3⁺ cells in the thymus of indicated mice. *, P < 0.05; **, P < 0.01; ***, P < 0.001. NS, not significant (P > 0.05).