A role for apoptosis-inducing factor in T cell development

Abstract

Apoptosis-inducing factor (Aif) is a mitochondrial flavoprotein that regulates cell metabolism and survival in many tissues. We report that aif-hypomorphic harlequin (Hq) mice show thymic hypocellularity and a cell-autonomous thymocyte developmental block associated with apoptosis at the β-selection stage, independent of T cell receptor β recombination. No abnormalities are observed in the B cell lineage. Transgenes encoding wild-type or DNA-binding-deficient mutant Aif rectify the thymic defect, but a transgene encoding oxidoreductase activity-deficient mutant Aif does not. The Hq thymic block is reversed in vivo by antioxidant treatment, and Hq T but not B lineage cells show enhanced oxidative stress. Thus, Aif, a ubiquitous protein, serves a lineage-specific nonredundant antiapoptotic role in the T cell lineage by regulating reactive oxygen species during thymic β-selection.

Figure 1.

Deficit of naive T cells in the peripheral lymphoid organs of Aif-hypomorphic Hq mice. (A) Representative two-color analyses of splenic and lymph node cells from 8–12-wk-old WT and Hq mice stained for CD4 and CD8, showing frequencies of cells in quadrants. (B–D) Frequencies (B) and absolute numbers of CD4 and CD8 cells in the spleen (C) and inguinal lymph node (D) of WT and Hq mice. *, P < 0.01. (E) Absolute numbers of B220⁺ B cells (B) and CD11b⁺ macrophages (M) in the spleen (S) and inguinal lymph node (LN) of WT and Hq mice. (F) Representative analysis of CD44 levels on gated CD4 or CD8 cells from spleen or lymph nodes of WT (thin lines) or Hq (thick lines) mice. Filled histograms represent isotype controls. (G) Absolute numbers of CD44^lo (naive) and CD44^hi (memory) CD4 and CD8 cells per organ in the spleen (S) and inguinal lymph nodes (LN) of WT and Hq mice. *, P < 0.01; **, P < 0.05. All data are shown as mean ± SE (n = 3–5) and are representative of at least three independent experiments.

Figure 2.

Thymus-specific T cell lineage–autonomous developmental blockade in Hq mice. (A) Representative analysis of thymocytes from 8–12-wk-old WT and Hq mice stained for CD4 and CD8, showing frequencies of cells in quadrants. Data are representative of three independent experiments. (B) Numbers of thymocyte subpopulations per thymus in WT and Hq mice, as mean ± SE (n = 4). *, P < 0.01. (C–F) Mixed bone marrow chimeric mice were constructed by giving CD45.1xCD45.2 F1 recipients WT CD45.1 bone marrow in a 1:1 ratio with CD45.2 WT or Hq marrow (as indicated). Chimeric mice were analyzed at 4–8 wk for the distribution of CD45 allotypes in various cell lineages in different tissues. Cells from chimeric mice were stained for donor CD45 allotypes and various lineage markers. Representative two-color plots are shown for peripheral blood cells stained for CD45 allotypes versus B220, CD11b, Gr-1, CD4, or CD8 (C), and bone marrow cells stained for CD45 allotypes versus B220 and Gr-1 and thymocytes stained for CD45 allotypes versus CD4 and CD8 (D). Frequencies of cells in each quadrant are indicated. Absolute numbers of lymph node CD4, CD8, B220, or CD11b cells (E) and of thymocyte DN, DP, CD4SP, and CD8SP subsets (F) of either CD45.1⁺ or CD45.2⁺ phenotype from the chimeric mice receiving either WT or Hq bone marrow as described. The groups shown are CD45.1 (WT donor) from chimeras given CD45.2 WT (CD45.1WT[WT]) or CD45.2 Hq (CD45.1WT[Hq]) marrow, and the CD45.2 cells from these chimeras (identified as CD45.2WT and CD45.2Hq). Data are shown as mean ± SE (n = 3). *, P < 0.01. Data are representative of five independent experiments.

Figure 3.

β-Selection stage thymic defect in Hq mice. (A) Representative analysis of CD44 and CD25 expression on gated DN thymocytes (negative for lineage markers CD3/CD4/CD8/B220/CD11b/CD11c/Gr-1) of 8–12-wk-old WT and Hq mice, showing frequencies of cells in quadrants. (B) Numbers of various thymocyte subsets per thymus in WT or Hq mice, from analyses as in A, shown as mean ± SE (n = 3). *, P < 0.01; **, P < 0.05. (C) Representative analysis of gated DN3 thymocytes from WT (thin line) or Hq (thick line) stained for CD27. Gray curve indicates isotype control. (D) Frequencies of CD27^lo and CD27^hi DN3 thymocytes from WT and Hq mice as indicated, from analyses as in C, shown as mean ± SE (n = 3). *, P < 0.01. (E–G) Magnetically sorted >90% pure DN1 + DN2 thymocytes from WT or Hq thymus were put in culture with the OP9DL1 stromal cell line. Cells were harvested and analyzed flow cytometrically on day 3 of culture. Two-color plots (E) show the expression profile of Thy-1 and CD44 on cells from WT or Hq cultures as indicated. Flow cytometric histograms (F) show the CD25 expression profile of gated Thy-1⁺CD44⁻ cells from WT (thin line) or Hq (thick line) cultures. Bar diagrams (G) show frequencies of DN3 (CD44⁻CD25⁺) and DN4 (CD44⁻CD25⁻) cells in the Thy-1⁺CD44⁻ compartment of WT or Hq cultures as mean ± SE (n = 3). *, P < 0.05. (H) B220 staining on bone marrow cells from WT (thin line) or Hq (thick line) mice. Filled histogram represents isotype control. (I) Representative two-color analysis of gated B220⁺ bone marrow cells from WT or Hq mice, stained for CD43 and IgM, showing frequencies of pro–B (CD43⁺IgM⁻), pre–B (CD43⁻IgM⁻), and B (CD43⁻IgM⁺) cells. (J) Numbers of B cell lineage subsets in bone marrow of WT or Hq mice, from analyses in H and I, shown as mean ± SE (n = 3). Data are representative of at least two independent experiments.

Figure 4.

Hq thymic defect is independent of TCR-β recombination and associated with cell death during β-selection. (A) Representative histogram overlay of gated DN3 thymocytes from WT (thin line) or Hq (thick line) mice, stained for intracellular TCR-β. Filled histogram represents isotype control. (B) Frequencies of intracellular TCR-β containing DN3 thymocytes in WT and Hq mice as calculated from the analysis shown in A. Data are shown as mean ± SE (n = 3). (C) Total thymocyte numbers in 4–6-wk-old F2 progeny littermate mice from Hq and DO11.10-TCR tg parentage. F1 mice generated from Hq × DO11.10 breeding were intercrossed. The resultant progeny were typed for the Hq allele and the TCR transgene and separated into four groups on that basis as shown. Data are shown as mean ± SE (n = 3–4). *, P < 0.01. (D) Representative histogram overlays of gated DN3 and DN4 thymocytes from WT (thin line) or Hq (thick line) mice given EdU in vivo and stained for EdU incorporation. Both EdU staining and cell size (FSC) are shown. Filled histograms represent isotype controls. (E) Frequencies of DN3 and DN4 thymocytes incorporating EdU in vivo in WT and Hq mice as calculated from the analysis shown in D. Data are shown as mean ± SE (n = 3). *, P < 0.01. (F) Annexin V staining on gated DN thymocytes from WT (thin line) or Hq (thick line) mice. Filled histogram represents isotype control. (G) Frequencies of annexin V–binding DN3 and DN4 thymocytes in WT and Hq mice. Data are shown as mean ± SE (n = 4). *, P < 0.01. (H) Frequencies of annexin V–binding DP thymocytes in WT and Hq mice. Data are shown as mean ± SE (n = 3). (I) DN1, DN3, and DN4 thymocytes from WT or Hq mice as indicated were electronically sorted and put in culture in vitro. After varying periods of time as shown, the frequencies of dead cells were estimated by Trypan blue staining. Data are shown as mean ± SE (n = 3). *, P < 0.05. Data are representative of at least three independent experiments.

Figure 5.

Restoration of normal thymic development and peripheral T cell phenotype in Hq mice by transgenically expressed Aif. (A) Plasmid map for the expression vector used for generating aif tg mice. (B) PCR results of screening for transgene inheritance on genomic DNA from ear clip tissue from the progeny of male mice receiving intratesticular injection of the plasmid shown in A. The vector control (+), a WT mouse DNA sample, and five sample progeny mouse samples are shown, along with a marker lane (M). (C) Western blot analysis of Aif expression in spleen cell lysates. Transgene-bearing male mice were bred with heterozygous Hq females to generate male littermates of all four genotypes, namely, non-tg WT, non-tg Hq, tg WT, and tg Hq, as indicated. (D) Representative two-color analyses of splenic cells from 6–8-wk-old WT and Hq mice of either non-tg or tg genotypes as indicated, stained for CD4 and CD8, showing frequencies of CD4 and CD8 cells. (E) Representative analysis of CD44 levels on gated CD4 or CD8 cells from spleen of WT (black lines) or Hq (red lines) mice of either non-tg (thin lines) or tg (thick lines) genotypes as shown. Filled histograms represent isotype controls. (F) Numbers of total thymocytes per thymus in 6–8-wk-old WT and Hq mice of either non-tg or tg genotypes as indicated, as mean ± SE (n = 3). *, P < 0.01. (G) Representative analysis of thymocytes from WT and Hq mice of either non-tg or tg genotypes as indicated, stained for CD4 and CD8, showing frequencies of subpopulations. (H) Representative analysis of CD44 and CD25 expression on gated DN thymocytes (negative for lineage markers CD3/CD4/CD8/B220/CD11b/CD11c/Gr-1) of WT and Hq mice of either non-tg or tg genotypes as indicated, showing frequencies of DN thymocyte subpopulations. (I–K) Frequencies of DN3 (I) and DN4 (J) thymocyte subsets, and mean size (as mean fluorescence intensity [MFI]) of DN4 thymocytes (K) from WT and Hq mice of either non-tg or tg genotypes as indicated, from analyses as in H, shown as mean ± SE (n = 3). *, P < 0.01. All flow cytometric analyses are representative of at least two independent experiments.

Figure 6.

Restoration of normal thymic development and peripheral T cell phenotype in Hq mice requires the oxidoreductase function but not the DNA-binding function of Aif. (A) Western blot analysis of Aif expression in spleen cell lysates. Mice bearing transgenes encoding the mutant Aif constructs Aif254/264 or Aif262/299 were bred with heterozygous Hq females to generate male tg WT and tg Hq littermates as shown, and spleen cell lysates were tested. (B–E) Frequencies of naive CD4 (B and D) and naive CD8 (C and E) cells in spleens from 6–8-wk-old WT and Hq littermate mice of either non-tg or tg genotypes (with either Aif254/264 or Aif262/299 as indicated) as shown, as mean ± SE (n = 3). *, P < 0.01. (F–I) Numbers per spleen of naive CD4 (F and H) and naive CD8 (G and I) cells in WT and Hq littermate mice of either non-tg or tg genotypes (with either Aif254/264 or Aif262/299 as indicated) as shown, as mean ± SE (n = 3). *, P < 0.01. (J and K) Numbers of thymocytes from 6–8-wk-old WT and Hq littermate mice of either non-tg or tg genotypes (with either Aif254/264 or Aif262/299 as indicated) as shown, as mean ± SE (n = 3). *, P < 0.01. (L–Q) Frequencies of DN (L and M), DN3 (N and O), and DN4 (P and Q) thymocytes from WT and Hq littermate mice of either non-tg or tg genotypes (with either Aif254/264 or Aif262/299 as indicated) as shown, as mean ± SE (n = 3). *, P < 0.01. All flow cytometric analyses are representative of at least two independent experiments.

Figure 7.

Increased ROS levels are involved in the thymic hypocellularity of Hq mice. (A) Representative two-color CD4 versus CD8 analysis of thymocytes from WT or Hq mice treated in vivo with Euk-134, as indicated. Frequencies of cells in each quadrant are shown. (B) Absolute numbers per thymus of total thymocytes and subsets as indicated in WT and Hq mice treated in vivo with Euk-134. Data are shown as mean ± SE (n = 3–4). *, P < 0.05. (C) Frequencies of DN thymocyte subsets as indicated in WT and Hq mice treated in vivo with Euk-134. Data are shown as mean ± SE (n = 3–4). Data are representative of three independent experiments.

Figure 8.

Increased ROS levels are involved in the thymic hypocellularity of Hq mice. (A–C) Representative histogram overlays of gated DN3 and DN4 thymocytes from WT (thin lines) or Hq (thick lines) mice stained with DCFDA (A), mBCl (B), or MitoSOX red (C) as indicated. Filled histograms represent isotype controls. (D) Representative histogram overlays of gated DP thymocytes from WT (thin lines) or Hq (thick lines) mice stained with DCFDA. Filled histogram represents negative control. (E) Representative histogram overlays of gated B220⁺CD43⁺IgM⁻ pro–B or B220⁺CD43⁻IgM⁻ pre–B cells from bone marrow of WT (thin lines) or Hq (thick lines) mice stained with DCFDA. Filled histograms represent negative controls. (F) Representative histogram overlays of gated DN3 and DN4 thymocytes from WT (thin lines) or Hq (thick lines) mice which were non-tg, Aif tg, Aif254/264 tg, or Aif262/299 tg as indicated, stained with the ROS indicator dye CellROX Deep Red. Filled histograms represent negative controls. Data represent three to eight independently tested littermate pairs. (G) DN3 thymocytes from WT or Hq mice as indicated were electronically sorted and put in culture in vitro for 8 h with (−) or without (+) Euk-134 as indicated, and the frequencies of dead cells were estimated by Trypan blue staining. Data are shown as mean ± SE (n = 3) *, P < 0.01. All data are representative of at least two independent experiments.