Helios marks strongly autoreactive CD4+ T cells in two major waves of thymic deletion distinguished by induction of PD-1 or NF-κB

Abstract

Acquisition of self-tolerance in the thymus requires T cells to discriminate strong versus weak T cell receptor binding by self-peptide-MHC complexes. We find this discrimination is reported by expression of the transcription factor Helios, which is induced during negative selection but decreases during positive selection. Helios and the proapoptotic protein Bim were coinduced in 55% of nascent CCR7(-) CD4(+) CD69(+) thymocytes. These were short-lived cells that up-regulated PD-1 and down-regulated CD4 and CD8 during Bim-dependent apoptosis. Helios and Bim were also coinduced at the subsequent CCR7(+) CD4(+) CD69(+) CD8(-) stage, and this second wave of Bim-dependent negative selection involved 20% of nascent cells. Unlike CCR7(-) counterparts, Helios(+) CCR7(+) CD4(+) cells mount a concurrent Card11- and c-Rel-dependent activation response that opposes Bim-mediated apoptosis. This "hollow" activation response consists of many NF-κB target genes but lacks key growth mediators like IL-2 and Myc, and the thymocytes were not induced to proliferate. These findings identify Helios as the first marker known to diverge during positive and negative selection of thymocytes and reveal the extent, stage, and molecular nature of two distinct waves of clonal deletion in the normal thymus.

Figure 1.

Helios expression differentiates thymocytes undergoing positive and negative selection. (A) Phenotype of Foxp3⁻ thymocytes from Bim-sufficient (Bim^+/+) or Bim-deficient (Bim^−/−) 3A9 TCR STg or 3A9 TCR × insHEL DTg mice (denoted far left) with gates defining the CD4⁺ 8⁺ (DP) and CD4⁺ 8⁻ (CD4^SP) populations (first column), which were analyzed for expression of TCR^3A9 versus CCR7 (second and third columns). Histograms (right) show labeling for Bim, Nur77, or Helios protein on Foxp3⁻ DP or CCR7⁺ TCR^3A9+ CD4^SP thymocytes. Data are representative of three experiments. (B) Number of CCR7⁺ TCR^3A9+ Foxp3⁻ CD4^SP thymocytes with at least seven mice per group compiled from four experiments. Error bars show SD. (C) Representative plots (left) show gating of Helios^+/− subsets of TCR^3A9 Foxp3⁻ CD4^SP thymocytes, which are enumerated in the graphs (right) with at least six Bim^+/+ or Bim^−/− STg or DTg mice per group compiled from four experiments. Numbers in plots indicate percentage of cells in gates shown, and columns in graphs show mean and SD. (D) Helios induction in mixed bone marrow chimeras with lower frequencies of TCR^3A9+ cells. Irradiated CD45.1⁺ insHEL^neg or insHEL^pos recipient mice were reconstituted with CD45.2⁺ Bim^−/− STg bone marrow cells alone or mixed at a ratio of 1:19 with CD45.1⁺ non-Tg marrow. Helios mean fluorescence intensity (MFI) on the Bim^−/− CD45.2⁺ CCR7⁺ TCR^3A9+ Foxp3⁻ CD4^SP thymocytes is plotted against their absolute number per thymus 4–5 wk after reconstitution. Helios MFIs were normalized to Helios MFI of non-Tg DP thymocytes analyzed in parallel. Data in D were compiled from two separate experiments; each symbol represents an individual chimera. Unpaired Student’s t test p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.

Figure 2.

Expanded Helios⁺ thymocyte populations in apoptosis-deficient mice. (A) CD24/CCR7 phenotype of non-Tg DP and CD4^SP thymocytes showing the gates used to define the CD4^SP subsets SP1 (top left), SP2 (top right), and SP3 (bottom right), which were examined for Qa-2 expression (middle); summary (right) shows the percentage of Qa-2⁺ cells within each CD4^SP subset compiled from two experiments. Error bars show SD. (B) Phenotypic progression within the CD4^SP stage. Non-Tg B6 mice were either left uninjected or injected i.p. with 1 mg BrdU either 1, 2, 3, 4, 5, 6, or 7 d before flow cytometry on thymocytes. Plots show labeling for TCRβ versus BrdU on all thymocytes (top row). BrdU⁺ thymocytes were analyzed for expression of CD4 versus CD8 (middle row), and BrdU⁺ CD4^SP cells were analyzed for CD24 versus CCR7 expression (bottom row). Data are representative of three experiments. (C) Helios⁺ thymocyte populations expand in apoptosis-deficient mice with natural TCR repertoires. Representative plots show phenotype of Foxp3⁻ thymocyte subsets in wild-type, Bim^−/−, and Vav-BCL2 Tg mice (denoted far left). The first column shows quadrant gates defining four populations on the basis of CD4 and CD8; in the third column, the CD4^SP population is further divided into the subsets SP1–SP4. The fourth through eighth columns display the Helios/Bim phenotype of populations denoted above the plots. Graphs (bottom) show number of Helios⁺ (left) or Helios⁻ (right) Foxp3⁻ thymocytes within each subpopulation as defined using gates represented in the plots above, compiled from eight separate experiments that used a total of 13 Bim^−/− and 8 B6.Vav-BCL2 Tg mice. Numbers in plots indicate percentage of cells in gates shown, and columns in graphs show mean and SD. The wild-type group comprises pooled data from 16 B10.BR mice and 10 B6 mice. No statistically significant differences existed between wild-type B10.BR and B6 mice. Paired (in A) or unpaired (in C) Student’s t test p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.

Figure 3.

Superantigen recognition by CCR7⁻ thymocytes induces Helios, PD-1, and acquisition of a CD4^lo 8^lo phenotype. (A) Each row displays thymocytes expressing TCR Vβ5, Vβ11, or Vβ8 from Bim^+/+ or Bim^−/− B10.BR mice (denoted far left). The gated subpopulations are denoted above the plots, and the markers displayed are below. Numbers in plots indicate percentage of cells in gates shown. Data are representative of five experiments. (B) Enumeration of Foxp3⁻ thymocytes expressing TCR Vβ5, Vβ11, or Vβ8 with a DP, CD4^lo 8^lo, CD4^SP, or CD8^SP phenotype as gated in A compiled from five experiments with a total of 9 Bim^−/− and 11 Bim^+/+ mice. Columns in graphs show mean and SD. Unpaired Student’s t test p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.

Figure 4.

Approximately 55% of TCR-signaled thymocytes die by Bim-dependent negative selection at the CCR7⁻ stage. (A) Analysis of nascent BrdU⁺ CCR7⁻ CD24⁺ thymocytes from wild-type mice 3 d after BrdU injection. Quadrants in top left panel define three major CD69 versus TCRβ subsets that are each analyzed for CD4 and CD8 expression on the right. Bottom panels reverse the gating order to show percentage of CD4^SP in the nascent pool and their expression of CD69 and TCRβ. Plots were concatenated from four samples, representative of two experiments each analyzing at least four mice individually. (B and C) Analysis of nascent BrdU⁺ CCR7⁻ CD24⁺ thymocytes concatenated from four Bim^+/+ or four Bim^−/− mice on B10.BR background 24 h (B) or 3 d (C) after BrdU treatment. Two Zap70^mrt/mrt mice were also analyzed to test TCR signaling. Top row: percentage of CD69⁺ cells. Second row: Helios and PD-1 expression on CD69⁺ subset. Bottom rows: CD4 and CD8 expression on Helios⁺ and Helios⁻ subsets of CD69⁺ cells. Percentage of gated cells is shown in the plots. Graphs show mean and SD percentage of CD69⁺ cells among BrdU⁺ CD24⁺ CCR7⁻ thymocytes from at least five mice per group compiled from two experiments in B and at least four mice per group from two experiments in C. Unpaired Student’s t test p-value: ***, P < 0.001.

Figure 5.

Organ-specific antigen in thymic medulla induces Helios and Bim in CCR7⁺ thymocytes, and Bim is required to prevent organ-specific CD4⁺ T cell peripheralization. (A) The first and second columns show the CD24/CCR7 phenotype of Foxp3⁻ DP and CD4^SP thymocytes in Bim^+/+ or Bim^−/− STg or DTg mice (denoted far left); the third through sixth columns show the Helios/Bim phenotype on the Foxp3⁻ thymocyte subsets DP–SP3. Plots are representative of three separate experiments. (B) Number of Foxp3⁻ thymocytes at each maturation stage from DP to SP3. (C) PD-1 is not induced in DTg thymocytes. Histograms show PD-1 expression on Foxp3⁻ DP and Foxp3⁻ CD4^SP thymocytes from Bim^+/+ or Bim^−/− DTg mice, representative of three experiments. (D) CD4^lo 8^lo thymocyte number is not increased in Bim^−/− DTg mice. CD4/CD8 phenotype (left) and number (right) of Foxp3⁻ TCR^3A9+ thymocyte subsets in Bim^+/+ or Bim^−/− STg or DTg mice are shown. (E) Frequency of TCR^3A9+ Foxp3⁻ CD4⁺ cells among splenic lymphocytes in Bim^+/+ or Bim^−/− STg or DTg mice. All column graphs show mean and SD for at least six STg mice and nine DTg mice per Bim genotype compiled from at least three experiments. Unpaired Student’s t test p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.

Figure 6.

A distinct hollow activation response accompanies Helios and Bim induction in strongly self-reactive thymocytes at the CCR7⁺ stage. (A) A microarray dataset (Liston et al., 2007) was mined for 53 mRNAs previously shown to be induced in primary human T cells activated for 3 h and 12 mRNAs associated with mitosis (Shaffer et al., 2001). 47 and 9 of these mRNAs, respectively, were represented on the array. For comparison, mRNAs encoding OX40 and GITR were added to this set. Sorted thymocytes analyzed by microarray were preselection CD4⁺ CD8⁺ CD69⁻ TCR^3A9− thymocytes from STg and DTg mice, CD4⁺ CD8^lo CD69⁺ TCR^3A9+ thymocytes from STg mice (undergoing positive selection), and CD4⁺ CD8^lo CD69⁺ TCR^3A9+ thymocytes from DTg mice (undergoing negative selection). T cell activation genes were categorized into three groups using the p-values of one-way ANOVA followed by Tukey’s multiple comparison tests (see Materials and methods). For each of the nine mitosis genes analyzed, the p-value of ANOVAs testing preselection with either positive or negative selection thymocytes was <0.05. Shading indicates fold change of the mean relative to preselection thymocytes. (B) Flow cytometry for proliferation antigen Ki-67 in CD4⁻ CD8⁻ (DN), CD4⁺ CD8⁺ (DP), and CCR7⁺ CD4^SP thymocytes from non-Tg, STg, and DTg mice, representative of two separate experiments. (C) Histogram overlays and column graphs compare expression of the indicated proteins on Helios⁺ (red) and Helios⁻ (black) Foxp3⁻ CD4^SP populations at SP1–SP3 maturation stages with corresponding DP cells (shaded histograms) from non-Tg mice. Column graphs below show the mean fluorescence intensity (MFI) for each marker on the indicated SP subsets relative to the MFI of Foxp3⁻ DP thymocytes from the same sample. Histograms in B and C are representative of at least three separate experiments; column graphs in C show mean ± SD for at least four mice compiled from at least two experiments. Paired Student’s t tests p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.

Figure 7.

Card11 and c-Rel mutations perturb Helios⁺ Foxp3⁻ CD4^SP thymocytes selectively at the SP2 and SP3 stages. (A) Efficient SP1 Helios⁺ Foxp3⁻ CD4^SP formation. Helios versus PD-1 expression on SP1 (CD24⁺ CCR7⁻) Foxp3⁻ CD4^SP thymocytes from wild-type, Card11^unm/unm, and Rel^−/− mice. (B) Reduced SP2 and SP3 Helios⁺ Foxp3⁻ CD4^SP formation in Card11^unm/unm and Rel^−/− mice. Helios versus Bim expression on SP2 (CD24⁺ CCR7⁺) and SP3 (CD24⁺ CCR7⁻) Foxp3⁻ CD4^SP thymocytes in the same set of samples shown in A. (C) Graphs show the percentage and number of Helios⁺ cells in the SP1, SP2, and SP3 Foxp3⁻ CD4^SP thymocyte populations for wild-type, Card11^unm/unm, and Rel^−/− mice on the B6 background with at least eight mice per group compiled from five separate experiments, two of which included mice of all three genotypes. (D) Plots (left) show Helios versus Bim expression on Foxp3⁻ CCR7⁺ CD4^SP (SP2 and SP3) thymocytes from non-Tg Card11^unm/unm mice (n = 5), age-matched wild-type B6 controls (n = 5), and Vav-BCL2 Tg Card11^+/+ (n = 4) or Card11^unm/unm (n = 5) littermates on the B6 background with a summary showing the mean and SD frequencies of Helios⁺ cells as gated in the plots, compiled from two separate experiments. (E) Card11^unm/unm or Bim^−/− mice and their wild-type littermates on the B10.BR background were injected i.p. with 1 mg BrdU 3 d before flow cytometry on thymocytes. Top panel shows CCR7⁻ CD24⁺ BrdU⁺ cells (concatenated from four to five mice per genotype) with the gate defining the Helios⁺ TCRβ⁺ population, which was analyzed for CD4/CD8 expression in the bottom panel. Graph (right) shows frequency of Helios⁺ TCRβ⁺ cells among CCR7⁻ CD24⁺ BrdU⁺ thymocytes. (F) Helios versus TCRβ expression (left) on nascent SP2 (CD24⁺ CCR7⁺ CD4^SP BrdU⁺) thymocytes in the same set of samples shown in E; graph (right) shows frequency of Helios⁺ TCRβ⁺ cells as gated on plots. Data in E and F represent four to five mice per group in a single experiment. Numbers in plots indicate percentage of cells in gates shown, and columns show mean and SD. Unpaired Student’s t test p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.

Figure 8.

A Card11-dependent response induced concurrently with Bim in self-reactive Foxp3⁻ CCR7⁺ CD4^SP thymocytes curbs clonal deletion and is rate limiting for thymic T reg cell differentiation. (A) Expression of the T cell activation markers CD25, OX40, GITR, and c-Rel versus Bim in Foxp3⁻ TCR^3A9+ CD4^SP thymocytes from STg or DTg mice of the indicated Card11 genotypes. Data are representative of four experiments with a total of at least nine mice of each Card11 genotype. (B and C) Reduced formation of mature phenotype Foxp3⁻ CD4^SP thymocytes in Card11^unm/unm DTg mice. Phenotype of Foxp3⁻ CD4^SP thymocytes from Card11^+/+ or Card11^unm/unm STg or DTg mice displaying Helios versus Bim (representative of two experiments; B) or TCR^3A9+ versus CCR7 (C). Summary (bottom) accompanies the data in C, showing the total number of cells in the top right quadrant (TCR^3A9+ CCR7⁺) per thymus, compiled from five experiments with at least 8 STg and 11 DTg mice per Card11 genotype. (D) Representative plots (top) show phenotype, and summary (bottom) shows number of TCR^3A9hi Foxp3⁺ CD4^SP thymocytes compiled from six experiments with at least 6 STg and 14 DTg mice per Card11 genotype. Numbers in plots indicate percentage of cells in gates shown, and summaries show mean and SD. Student’s t tests p-values: ***, P < 0.001; **, P = 0.001–0.01; *, P = 0.01–0.05.