The thymic medulla is required for Foxp3+ regulatory but not conventional CD4+ thymocyte development

Abstract

A key role of the thymic medulla is to negatively select autoreactive CD4(+) and CD8(+) thymocytes, a process important for T cell tolerance induction. However, the involvement of the thymic medulla in other aspects of αβ T cell development, including the generation of Foxp3(+) natural regulatory T cells (nTreg cells) and the continued maturation of positively selected conventional αβ T cells, is unclear. We show that newly generated conventional CD69(+)Qa2(-) CD4 single-positive thymocytes mature to the late CD69(-)Qa2(+) stage in the absence of RelB-dependent medullary thymic epithelial cells (mTECs). Furthermore, an increasing ability to continue maturation extrathymically is observed within the CD69(+)CCR7(-/lo)CCR9(+) subset of conventional SP4 thymocytes, providing evidence for an independence from medullary support by the earliest stages after positive selection. In contrast, Foxp3(+) nTreg cell development is medullary dependent, with mTECs fostering the generation of Foxp3(-)CD25(+) nTreg cell precursors at the CD69(+)CCR7(+)CCR9(-) stage. Our results demonstrate a differential requirement for the thymic medulla in relation to CD4 conventional and Foxp3(+) thymocyte lineages, in which an intact mTEC compartment is a prerequisite for Foxp3(+) nTreg cell development through the generation of Foxp3(-)CD25(+) nTreg cell precursors.

Figure 1.

CCR7 and CCR9 define distinct subsets of SP4 thymocytes. (A) CD69⁺ and CD69⁻ subsets of SP4 thymocytes from Rag2GFP mice analyzed for CCR7/CCR9 expression. Data are typical of four experiments. (B) Levels of HSA, CD62L, Qa2, and Rag2GFP in the following SP4 subsets: CD69⁺CCR7^−/loCCR9⁺ (red), CD69⁺CCR7⁺CCR9⁻ (blue), and CD69⁻CCR7⁺CCR9⁻ (green). For comparison, Rag2GFP expression by CD69⁻CD4⁺8⁺ thymocytes is shown (black). Data are typical of three separate experiments.

Figure 2.

Developmental emergence of conventional and Foxp3⁺ T_reg SP4 cells. (A) qPCR in indicated thymocyte subsets. Error bars indicate SEM. mRNA levels were normalized to β-actin. Data are from at least two independently sorted biological samples, with each gene analyzed two times. (B) Phenotype of purified CD69⁺CCR7^−/loCCR9⁺ SP4 thymocytes before (left) and after (right) incorporation into reaggregate thymus organ cultures (RTOCs). Data are typical of four experiments. (C) Developmental sequence of SP4/nT_reg thymocyte maturation. (D) CCR7/CCR9 subsets of adult WT αβ TCR^hi SP4 thymocytes for CD69, CD25, and Foxp3 expression. Data represent at least three separate experiments.

Figure 3.

RelB-dependent mTECs are dispensable for conventional SP4 thymocyte development. (A) Qa2/CD69 expression in WT (top) and Relb^−/− (bottom) thymocytes after gating on αβ TCR^hi SP4 thymocytes. Data are representative of three experimental replicates. (B) Immunofluorescent staining of WT and Relb^−/− graft sections for ERTR5 and β5t. C denotes cortex, and M denotes medulla. Bars: (left) 200 µm; (right) 100 µm. (C) Thymocytes from WT (top) and Relb^−/− (bottom) thymus grafts, with Qa2/CD69 levels shown for αβ TCR^hi SP4 cells. (D) Frequencies of Qa2/CD69 SP4 thymocytes subsets recovered from WT and Relb^−/− grafts. (E) Qa2/HSA/CD44 expression in SP4 thymocytes recovered from WT thymus (top), WT TEC grafts (middle), and Relb^−/− TEC grafts (bottom). (F) Mean fluorescence intensity (MFI) of Qa2 expression in CD69⁻ SP4 T cells from WT spleen or CD69⁻ SP4 thymocytes from WT or Relb^−/− TEC grafts. Error bars represent SEM; data in C–F are from at least three independent experiments, with a minimum of five of each graft type per experiment. In an unpaired Student’s two-tailed t test, ns denotes a nonsignificant difference where P > 0.1; ***, P < 0.001.

Figure 4.

Extrathymic development of CCR7^−/loCCR9⁺CD69⁺ SP4 thymocytes. (A) Autoantibodies in serum from nude mice receiving WT or Relb^−/− TEC grafts (top) and histological analysis (bottom) of lymphocytic infiltrates (arrows) in liver of the same mice. Bars, 100 µm. (B) CD44/CD62L expression in SP4 LN T cells from nude mice receiving either WT or Relb^−/− TEC grafts. Right panels show intracellular IFN-γ in CD3⁺ LN T cells of the same mice. Data in A and B are typical of three experimental replicates. (C) HSA/CD62L/Qa2/CD69 expression in CD69⁺CCR7^−/loCCR9⁺ adult αβ TCR^hi SP4 thymocytes before transfer (input) and their SP4 T cell progeny (CD45.2⁺ cells) recovered from both spleen and LN after 7 d (red lines). Black lines show host CD45.1⁺ SP4 LN and splenic T cells for comparison. (D) Ratio of recovered SP4 progeny in spleen 7 d after coinjection of equal numbers of either CD69⁺CCR7^−/loCCR9⁺ SP4 thymocytes and CD4⁺8⁺69⁺ thymocytes (left) or CD69⁺CCR7^−/loCCR9⁺ SP4 thymocytes and CD69⁻CCR7⁺ SP4 thymocytes (right). Error bars represent SEM, and data in C and D are typical of four separate experiments.

Figure 5.

RelB-dependent mTECs control Foxp3⁻CD25⁺ nT_reg cell precursor generation. (A) Expression of Foxp3GFP/CD25 in the LN progeny of i.v. injected CD69⁺Foxp3⁻CD25⁺ (left) and CD69⁺CCR7^−/loCCR9⁺ (right) T cells after gating on CD45.2⁺ injected cells. Data are typical of two experimental replicates. (B and C) WT (B) and Relb^−/− (C) thymus grafts analyzed for Foxp3⁻CD25⁺ nT_reg cell precursors and Foxp3⁺ nT_reg cells within the SP4 subset. (D) Quantitation of these populations; error bars represent SEM. An unpaired Student’s two-tailed t test was performed: ***, P < 0.001; **, P < 0.01. Data in B–D represent four separate experiments. (E) Confocal analysis of WT and Relb^−/− grafts. Bars, 100 µm.