Identification of a genomic enhancer that enforces proper apoptosis induction in thymic negative selection

Abstract

During thymic negative selection, autoreactive thymocytes carrying T cell receptor (TCR) with overtly strong affinity to self-MHC/self-peptide are removed by Bim-dependent apoptosis, but how Bim is specifically regulated to link TCR activation and apoptosis induction is unclear. Here we identify a murine T cell-specific genomic enhancer E^{BAB (Bub1-Acoxl-Bim)}, whose deletion leads to accumulation of thymocytes expressing high affinity TCRs. Consistently, E^BAB knockout mice have defective negative selection and fail to delete autoreactive thymocytes in various settings, with this defect accompanied by reduced Bim expression and apoptosis induction. By contrast, E^BAB is dispensable for maintaining peripheral T cell homeostasis via Bim-dependent pathways. Our data thus implicate E^BAB as an important, developmental stage-specific regulator of Bim expression and apoptosis induction to enforce thymic negative selection and suppress autoimmunity. Our study unravels a part of genomic enhancer codes that underlie complex and context-dependent gene regulation in TCR signaling.

Fig. 1

Identification of a T cell-specific cis-regulatory element E^BAB. a ChIP-seq visualization of H3K27ac in several mouse tissues. H3K27ac profiles from the thymus, spleen, bone marrow, liver, kidney, heart, brain, testis, and brown adipose tissues (BAT) are visualized using the UCSC genome browser (mm9). The E^BAB region is highlighted. b DNase hypersensitivity sites (DHS) in the same locus shown in (a). DHS profiles from the thymus, T-Naïve CD4⁺, regulatory T (T_reg) cells, spleen, and B cells (CD43⁻ or CD19⁺) are visualized using the UCSC genome browser (mm9). The E^BAB region is highlighted. c Schematic representation of ΔE^BAB. Arrowheads indicate the primers listed in Supplementary Data 1. d Genomic PCR against the E^BAB locus of WT, heterozygotes (E^BAB+/−), and ΔE^BAB mice. A representative gel-image of founder #44-derived DNAs is shown. See also Supplementary Fig. 2 for the results from #47- and #50-derived DNAs. e, f qPCR analysis for Bim (e) and Bub1 (f) on thymocytes, splenocytes, lung, liver, kidney, and pancreas. Data are pooled from five independent experiments (thymocyte, splenocyte; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 7–17 weeks old, mean ± s.d.) or three independent experiments (lung, liver, kidney, pancreas; n = 3 WT & E^BAB+/−–ΔE^BAB littermate pairs, 10–17 weeks old). Each symbol represents an individual mouse; small horizontal lines indicate the mean. No statistically significant differences between WT and E^BAB+/− and ΔE^BAB were detected (P ≥ 0.05; unpaired two-tailed Student’s t test)

Fig. 2

The ΔE^BAB thymus accumulates high affinity TCR clones. a Flow cytometric analysis of CD4 versus CD8 thymocyte populations. Data are representative of seven independent experiments (Young; n = 7 WT and E^BAB+/−–ΔE^BAB littermate pairs, 7–17 weeks old) or five independent experiments (Aged; n = 5 WT & E^BAB+/−–ΔE^BAB littermate pairs, 30–36 weeks old). b–d Double negative (DN), double positive (DP), CD4 single positive (SP), and CD8 SP thymocyte proportions of young and aged mice (b), and cell numbers of young (c) and aged (d) mice. Data are pooled from seven independent experiments (Young; n = 7 WT and E^BAB+/−–ΔE^BAB littermate pairs, 7–17 weeks old) or five independent experiments (Aged; n = 5 WT and E^BAB+/-–ΔE^BAB littermate pairs, 30–36 weeks old). e Scatter plot (log₂ fold change versus –log₁₀ P value) of genes analyzed by RNA-seq (n = 2 WT and E^BAB+/−–ΔE^BAB littermate pairs, 10–11 weeks old). Genes showing more than twofold changes with P < 0.05 (unpaired two-tailed Student’s t test) are highlighted. f Flow cytometric analysis of TCRβ versus CD69 thymocyte populations. Data are representative of five independent experiments (Young; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 9–12 weeks old, Aged; n = 6 WT and E^BAB+/− mice and n = 7 ΔE^BAB mice from WT and E^BAB+/−–ΔE^BAB littermate pairs and trios, 30–36 weeks old). g Post-selection (TCRβ^highCD69^high) thymocyte proportion of young and aged mice. Data are pooled from five independent experiments (Young; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 9–12 weeks old, Aged; n = 6 WT and E^BAB+/− mice, n = 7 ΔE^BAB mice, WT and E^BAB+/−–ΔE^BAB littermate pair or trio, 30–36 weeks old). h qPCR analysis for Nr4a1 in pre-selection (TCRβ^lowCD69^low) and post-selection thymocytes of WT and ΔE^BAB. Data are pooled from five independent experiments (n = 5 sex-matched WT–ΔE^BAB pairs, 10–18 weeks old, mean ± s.d.). Edges of the box are the 25th and 75th percentiles, and error bars extend to the maximum and minimum. Each symbol in (b, c, d, g) represents an individual mouse; small horizontal lines indicate the mean. n.s. not significant (P ≥ 0.05); *P < 0.05, **P < 0.01, ***P < 0.001 (unpaired two-tailed Student’s t test)

Fig. 3

E^BAB contributes to depletion of high affinity TCR clones in ex vivo TCR stimulation and the HY tg models. a Annexin V⁺Propidium Iodide (PI)⁻ fraction of total thymocytes stimulated with anti-CD3 and anti-CD28 antibodies (10 µg/ml) for 9 h. The percentage in mock sample (i.e., background) is subtracted. Data are pooled from 11 independent experiments (n = 3 WT–ΔE^BAB littermate pairs, n = 3 sex-matched WT–ΔE^BAB pairs, n = 1 WT–Bim^−/− littermate pair, n = 1 sex-matched WT–Bim^−/− pair, n = 3 sex-matched WT–Bim^−/−–ΔE^BAB trios, 5–20 weeks old). Edges of the box are the 25th and 75th percentiles, and error bars extend to the maximum and minimum. Outliers are defined as the data point that is located outside of q₃ + 1.5(q₃ − q₁) and q₁ − 1.5(q₃ − q₁), in which q₁ and q₃ are the 25th and 75th percentiles. b CD4 versus CD8 flow cytometric analysis of PI⁻HY-TCR⁺ thymocytes from female and male HY tg mice. The number in the plot is representative percentage of each gate. c−d DP thymocyte proportion (c) and cell numbers (d) of PI⁻HY-TCR⁺ thymocytes from female and male HY tg mice. e, f CD8 SP thymocyte proportion (e) and cell numbers (f) of PI⁻HY-TCR⁺ thymocytes from female and male HY tg mice. Data are representative of (b), or pooled from (c–f), six independent experiments (n = 1 Bim^+/−; HY⁺ and Bim^−/−; HY⁺ female mice, n = 5 Bim^+/−; HY⁺ and Bim^−/−; HY⁺ male mice, n = 3 E^BAB+/−; HY⁺ and ΔE^BAB; HY⁺ female mice, n = 5 E^BAB+/−; HY⁺ and ΔE^BAB; HY⁺ male mice, 6–8 weeks old). Each symbol in (c–f) represents an individual mouse; small horizontal lines indicate the means. n.s. not significant (P ≥ 0.05); *P < 0.05, **P < 0.01 (unpaired one-tailed Student’s t test or Mann–Whitney U test for Bim KO data, and unpaired two-tailed Student’s t test for ΔE^BAB data)

Fig. 4

E^BAB is essential for depleting high affinity TCR clones in the OT-II tg model. a CD4 versus CD8 flow cytometric analysis of TCR Vβ5⁺ thymocytes from mice injected with OVA_257–264 or OVA_323–229 peptide intraperitoneally. The number in the plot is representative percentage of each gate. b CD4 SP thymocyte proportion of TCR Vβ5⁺ thymocytes from mice treated as in Fig. 4a. c TCRβ versus CD69 flow cytometric analysis of TCR Vβ5⁺ thymocytes from mice treated as in Fig. 4a. The number in the plot is representative percentage of the gate. d Post-selection (TCRβ^highCD69^high) thymocyte proportion of TCR Vβ5⁺ thymocytes from mice treated as in Fig. 4a. Data are representative of (a, c) or pooled from (b, d) 13 independent experiments (n = 5 WT and E^BAB+/− mice injected with OVA_257–264 peptide, n = 8 WT and E^BAB+/−; OT-II⁺ mice injected with OVA_323–339 peptide, n = 6 ΔE^BAB; OT-II⁺ mice injected with OVA_257–264 and OVA_323–339 peptide, n = 3 Bim^+/−; OT-II⁺ and Bim^−/−; OT-II⁺ mice injected with each peptide, 5–14 weeks old). Each symbol (b, d) represents an individual mouse; small horizontal lines indicate the mean. n.s. not significant (P ≥ 0.05); **P < 0.01, ***P < 0.001 (unpaired one-tailed Student’s t test for Bim KO data, unpaired two-tailed Student’s t test for ΔE^BAB data)

Fig. 5

E^BAB regulates Bim expression upon TCR stimulation. a–d Gene expression changes for Bim (a), Nr4a1 (b), Bub1 (c), and Bcl2 (d) in total thymocytes stimulated for 3 h with anti-CD3 and anti-CD28 antibodies (10 µg/ml). e, f Gene expression level (e) and relative gene expression to WT (f) for Bim in pre-selection (TCRβ^lowCD69^low) and post-selection (TCRβ^highCD69^high) thymocytes. g, h Gene expression level (g) and relative gene expression to WT (h) for Bub1 in pre-selection (TCRβ^lowCD69^low) and post-selection (TCRβ^highCD69^high) thymocytes. Data are pooled from six independent experiments (a–d; n = 6 sex-matched WT–ΔE^BAB pairs, 9–18 weeks old) or five independent experiments (e–h; n = 5 sex-matched WT–ΔE^BAB pairs, 10–18 weeks old). Edges of the box are the 25th and 75th percentiles, and error bars extend to the maximum and minimum. Outliers are defined as the data point that is located outside of q₃ + 1.5(q₃ − q₁) and q₁ − 1.5(q3 − q₁), in which q₁ and q₃ are the 25th and 75th percentiles. n.s. not significant (P ≥ 0.05); *P < 0.05, **P < 0.01, ***P < 0.001 (unpaired two-tailed Student’s t test)

Fig. 6

E^BAB does not play a major role in maintaining homeostasis of regulatory T cells and peripheral T cells. a, b CD4⁺Foxp3⁺ thymic (a) and splenic (b) T_reg cell numbers of young and aged mice. Data are pooled from five independent experiments (Young; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 9–12 weeks old, Aged; n = 5 WT E^BAB+/−–ΔE^BAB littermate pairs, 30–36 weeks old). c Flow cytometric analysis of CD4 versus CD8 T cell populations in the spleen. The number in the plot is representative percentage of each gate. Data are representative of five independent experiments (Young; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 7–17 weeks old, Aged; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 30–36 weeks old). d CD4⁺ and CD8⁺ splenic T cell numbers. Data are pooled from five independent experiments (Young; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 7–17 weeks old, Aged; n = 5 WT and E^BAB+/−–ΔE^BAB littermate pairs, 30–36 weeks old). Each symbol in (a, b, d) represents an individual mouse; small horizontal lines indicate the mean. n.s. not significant (P ≥ 0.05) (unpaired two-tailed Student’s t test)

Fig. 7

E^BAB is dispensable for two apoptotic pathways in peripheral T cells. a, b Viability (% Annexin V⁻PI⁻) of activated CD4⁺ (a) and CD8⁺ (b) splenic T cells after IL-2 withdrawal. Data are pooled from ten independent experiments (a; n = 5 sex-matched WT–ΔE^BAB pairs and n = 5 sex-matched WT–Bim^−/−–ΔE^BAB trios, 5–26 weeks old, mean ± s.d.) or 11 independent experiments (b; n = 5 sex-matched WT–ΔE^BAB pairs and n = 5 sex-matched WT–Bim^−/−–ΔE^BAB trios, 5–26 weeks old, mean ± s.d.). c, d Gene expression level for Bim in activated CD4⁺ (c) and CD8⁺ (d) splenic T cells cultured with or without IL-2 for 6 h. Data are pooled from five independent experiments (n = 5 sex-matched WT–Bim^−/−–ΔE^BAB trios, 5–14 weeks old). e, f Expression of Bub1 in activated CD4⁺ (e) and CD8⁺ (f) splenic T cells cultured with or without IL-2 for 6 h. Data are pooled from five independent experiments (n = 5 sex-matched WT–Bim^−/−–ΔE^BAB trios, 5–14 weeks old). g, h Apoptotic cell rate (%Annexin V⁺) of activated CD4⁺ (g) and CD8⁺ (h) splenic T cells restimulated with anti-CD3 antibody (5 µg/ml) for 6 h. Data are pooled from five independent experiments (n = 5 sex-matched WT–Bim^−/−–ΔE^BAB trios, 5–26 weeks old). Edges of the box are the 25th and 75th percentiles, and error bars extend to the maximum and minimum. Outliers are defined as the data point that is located outside of q₃ + 1.5(q3 − q₁) and q₁ − 1.5(q₃ − q₁), in which q₁ and q₃ are the 25th and 75th percentiles (c–h). n.s. not significant (P ≥ 0.05); *P < 0.05, **P < 0.01, ***P < 0.001 (unpaired two-tailed Student’s t test)

Fig. 8

ΔE^BAB mice do not show any autoimmune phenotype. a Flow cytometric analysis of CD62L versus CD44 populations in CD4⁺ or CD8⁺ splenic T cell. The number in the plot is representative percentage of each gate. For CD4⁺ T cells, gates indicate activated (CD44^highCD62L^low) and naïve (CD44^lowCD62L^high) populations. For CD8⁺ T cells, activated effector memory (CD44^highCD62L^low), activated central memory (CD44^highCD62L^high) and naïve (CD44^lowCD62L^high) populations are gated. Data are representative of four independent experiments (n = 5 for WT and E^BAB+/− mice, n = 6 for ΔE^BAB mice from WT & E^BAB+/−–ΔE^BAB littermate pairs and trios, 30–34 weeks old). b, c Activated CD4⁺ (b) and CD8⁺ (c) proportions of aged mice. Data are pooled from four independent experiments (n = 5 for WT and E^BAB+/− mice, n = 6 for ΔE^BAB mice from WT and E^BAB+/−–ΔE^BAB littermate pairs and trios, 30–34 weeks old). d The mean (±s.e.m.) clinical scores at the days after EAE was induced in WT (control) (n = 10) and ΔE^BAB mice (n = 8). The incidence of EAE: control 10/10, ΔE^BAB 8/8. No data point showed statistically significant difference between WT and ΔE^BAB (i.e., unpaired two-tailed Student’s t test P ≥ 0.05). e Representative pictures for Hematoxylin and Eosin staining for the lung, liver, kidney, and pancreas of two independent experiments (n = 2 WT and E^BAB+/−–ΔE^BAB littermate pairs, 30–31 weeks old). The scale bars represent 100 μm. Each symbol in (b, c) represents an individual mouse; small horizontal lines indicate the mean. n.s. not significant (P ≥ 0.05) (unpaired two-tailed Student’s t test)

Fig. 9

Graphical summary. E^BAB plays a role in inducing TCR signal-dependent apoptosis to eliminate high affinity TCR clones in the thymus while dispensable for Bim-dependent peripheral T cell tolerance