Clonal deletion of thymocytes can occur in the cortex with no involvement of the medulla

Abstract

The thymic medulla is generally held to be a specialized environment for negative selection. However, many self-reactive thymocytes first encounter ubiquitous self-antigens in the cortex. Cortical epithelial cells are vital for positive selection, but whether such cells can also promote negative selection is controversial. We used the HY(cd4) model, where T cell receptor for antigen (TCR) expression is appropriately timed and a ubiquitous self-antigen drives clonal deletion in male mice. We demonstrated unambiguously that this deletion event occurs in the thymic cortex. However, the kinetics in vivo indicated that apoptosis was activated asynchronously relative to TCR activation. We found that radioresistant antigen-presenting cells and, specifically, cortical epithelial cells do not efficiently induce apoptosis, although they do cause TCR activation. Rather, thymocytes undergoing clonal deletion were preferentially associated with rare CD11c(+) cortical dendritic cells, and elimination of such cells impaired deletion.

Figure 1.

The phenotype of thymocytes undergoing positive and negative selection in the HY^cd4 model. (A) CD4 by CD8 profiles of T3.70⁺ thymocytes from female and male mice. (B) The expression of CD69 and PD-1 on thymocytes of the indicated genotype. The percentage of PD-1⁺ cells in the HY^cd4 male is indicated. (C) The activation of Caspase 3 in male mice compared with female mice. The fold change in male over female is indicated. Data represent the mean from 16 individuals from 12 different experiments ± SD. P < 0.0001.

Figure 2.

Apoptosis occurs asynchronously in vivo, with some cells surviving up to 4 d. BrdU was injected i.p. at the indicated number of hours before harvest. (A) The expression of CD4 and CD8 on BrdU⁺ T3.70⁺ thymocytes from representative mice from the indicated time point. (B) Expression of CD5, CD69, and PD-1 on BrdU⁺ T3.70⁺ thymocytes over time. (C) The total number of BrdU⁺ T3.70⁺ thymocytes versus time after BrdU injection. Data represent the mean ± SD from four separate time courses including three to seven individual mice. *, P = 0.0118 at 48 h, P = 0.0070 at 72 h, and P = 0.0037 at 96 h. (D) The frequency of active Caspase 3⁺ cells as a percentage of BrdU⁺ T3.70⁺ thymocytes. Data represent the mean ± SD from four separate time courses including three to seven individual mice. *, P = 0.0002 at 12 h, P < 0.0001 at 24 h, P = 0.0004 at 48 h, and P = 0.0061 at 72 h. Activation of Caspase 3 in vitro was determined by stimulating female HY^cd4 thymocytes with HYp peptide plus spleen APC for 8 h. Data were normalized for nonspecific death caused by in vitro culture and represent the mean ± SD from triplicate wells.

Figure 3.

HY^cd4 mixed BM chimeras as a model system to study clonal deletion in vivo. (A) HY^cd4 TCRα^o female BM was mixed with congenic B6.PL female and male BM and transferred into congenic female and male hosts. Chimeras were harvested at 6 wk and thymi were analyzed by flow cytometry for expression of Thy1.2, T3.70, CD4, CD8, and active Caspase 3. HY^cd4 donor-derived cells were identified by Thy1.2 where indicated. Numbers represent frequency of parent gate. (B) The activation of Caspase 3 in male chimeras compared with female chimeras. The fold change in male over female is indicated. Data represent the mean from 15–17 individuals from six different experiments ± SD. P < 0.0001.

Figure 4.

CCR7 is dispensable for clonal deletion of HY^cd4 thymocytes. (A) The expression of CCR7 on female and male T3.70⁺ thymocytes from HY^cd4 mice on a WT or Bim^o background. (B) BM from female HY^cd4 mice on either WT or CCR7° background was mixed with WT female or male BM and transferred into WT female or male recipients. Activation of Caspase 3 in T3.70⁺ thymocytes of the indicated genotype is depicted. Data represent the mean from four individuals from one experiment ± SD. P = 0.0002 when comparing either WT males and females or CCR7° males and females. The difference between WT males and CCR7° males was not significant. (C) HY^cd4 D^bo BM was mixed with WT or plt/plt female or male BM and transferred into WT or plt/plt female or male mice. The expression of CD69 and PD-1 on T3.70⁺ thymocytes from the indicated group is shown. (D) The activation of Caspase 3 on T3.70⁺ thymocytes. Data depicts the mean ± SD. The fold change is indicated. P = 0.0066. Data are representative of three separate experiments.

Figure 5.

Colocalization of active Caspase 3 and Thy1.2 identifies male-reactive thymocytes undergoing clonal deletion throughout the cortex. Thymus tissue sections from mice in Fig. 3 were analyzed by immunofluorescence for Thy1.2, active Caspase 3, and G8.8. (A) Images were overlayed and examined for colocalization between Thy1.2 and active Caspase 3. White lines are drawn around G8.8⁺ cells to identify the medulla and boxes identify colocalized cells. C, cortex; M, medulla. Bar, 250 μm. (B) Colocalization was confirmed at high magnification. Bar, 5 μm. (C) The frequency of colocalization is expressed as the number of colocalized cells per total Thy1.2⁺ cells in ten different images per thymus. Data represent the mean from four different males ± SD. (D) The relative distance of colocalized cells from the medulla was digitally calculated using Photoshop. The drawn line indicates the median of the dataset. The entire figure is representative of two individual groups of chimeras and three replicates of immunofluorescence staining.

Figure 6.

Radioresistant cells and cortical epithelial cells inefficiently induce apoptosis. HY^cd4 D^bo female BM was mixed with female or male BM from WT mice and transferred into WT female and male recipients as controls. (A) A cohort of WT male mice reconstituted with a mixture of HY^cd4 D^bo female BM and nontransgenic D^bo female BM was added. Data are representative of two separate experiments. (B) A cohort of female mice expressing the K14-HYp transgene reconstituted with a mixture of HY^cd4 D^bo BM and WT female BM was added. Data are representative of five separate experiments. The frequency of active Caspase 3⁺ T3.70⁺ cells in chimeric mice is shown. Data represent the mean from three to six individuals ± SD. Fold change over the female control group is indicated. A, P = 0.0463 when comparing the male radioresistant group to the WT female group and P = 0.1191 when comparing the male radioresistant group to the WT male group; B, P = 0.0042 when comparing WT male and female groups and P = 0.0085 when comparing the WT male group to the female K14-HYp group.

Figure 7.

Male-reactive cells undergoing apoptosis preferentially associate with cortical DCs. (A) Immunofluorescence staining for CD11c in the thymus. Areas of cortex and medulla are indicated. Bar, 250 μm. (B) Thymus tissue sections from mice in Figs. 3 and 5 were analyzed by immunofluorescence for Thy1.2, active Caspase 3, and CD11c. A Thy1.2 and active Caspase 3⁺ colocalized cell immediately adjacent to a CD11c⁺ cell is indicated by arrows. Bar, 5 μm. (C) The frequency of cells associated with CD11c⁺ cells was quantified for the indicated subsets. Data represent the mean from four different males ± SD. P < 0.0001. The frequency of cells associated with F4/80⁺ cells is also depicted. Data are representative of two individual groups of chimeras and one set of immunofluorescence staining.

Figure 8.

Cortical DCs are efficient APC for inducing clonal deletion. Female HY^cd4 D^bo BM was mixed with nontransgenic D^bo BM or CD11c-DTR female BM and transferred into D^bo or WT female recipients, respectively, as controls. Female HY^cd4 D^bo BM was also mixed with male CD11c-DTR BM and transferred into WT females or males and either left untreated or treated with DTx. Activation of Caspase 3 in treated versus untreated T3.70⁺ thymocytes of the indicated group is depicted. Data represent the mean from three to five mice ± SD. The fold change comparing untreated and DTx-treated mice for each group is indicated. P < 0.0001 for the male into female group and P = 0.0242 for the male into male group. ND, not done. Data are representative of four separate experiments.