The action of Bax and bcl-2 on T cell selection

Abstract

T cell development and selection in the thymus are shaped by the induction of apoptosis. However, a direct role in T cell development and selection for any of the molecules known to regulate apoptosis has remained controversial. We have studied the effect of bax and bcl-2 transgenes in recombination activation gene 1-deficient (RAG-1(-/-)) mice transgenic for the major histocompatibility complex class I-restricted F5 T cell receptor. Overexpression of a bax transgene in the thymus seriously impairs the production of mature T cells, whereas bcl-2 overexpression greatly promotes it. The effect of bax and bcl-2 overexpression on antigen-induced negative selection was studied using fetal thymic organ cultures. This analysis showed that Bcl-2 strongly inhibits negative selection, whereas Bax does not affect it. Our data directly show that Bcl-2 family members have specific roles in T cell selection and also lend support to the hypothesis that Bax and Bcl-2 can antagonize each other's action in a certain apoptosis pathway while in another they can be functionally nonreciprocal.

Figure 1

Flow cytometric analysis of T cell development in the thymus of F5/RAG-1^−/−/bax H-2^b/b mice and littermate controls lacking the bax transgene. Thymocytes were isolated from F5/RAG-1^−/−/bax H-2^b/b mice with or without the bax transgene then stained for CD4 (PE), CD8 (FITC), and either V_β11 or CD69 (both biotinylated). (A) Dot plots of CD4- and CD8-stained thymocytes with gates on the CD4^hiCD8^hi, CD4^loCD8^hi, and CD4⁻CD8^hi populations. The numbers represent the percentage of the total number of thymocytes in each gate. (B) Histograms represent V_β11 or CD69 expression on the thymocytes in each gate. The shaded region represents F5/RAG-1^−/− mice, whereas the black line represents the same mice plus the bax transgene.

Figure 2

Flow cytometric analysis of T cell development in the thymus of F5/RAG-1^−/−/bcl-2 H-2^b/b mice and littermate controls lacking the bcl-2 transgene. Thymocytes were isolated from F5/RAG-1^−/−/bcl-2 H-2^b/b mice with or without the bcl-2 transgene then stained for CD4, CD8, and either V_β11 or CD69. (A) Dot plots of CD4- and CD8-stained thymocytes with gates on the CD4^hiCD8^hi, CD4^loCD8^hi, and CD4⁻ CD8^hi populations. The numbers represent the percentage of the total number of thymocytes in each gate. (B) Histograms represent V_β11 or CD69 expression on the thymocytes in each gate. The shaded region represents F5/RAG-1^−/− mice, whereas the black line represents the same mice plus the bcl-2 transgene.

Figure 3

Direct measurement of the level of apoptosis caused by antigen induced negative selection in fetal thymic organ culture. FTOCs from F5/RAG-1^−/− mice with or without the bax or bcl-2 transgene were cultured for 4 d in vitro. 10 μM NP68 peptide, the cognate antigen for the F5 TCR, was then added to the culture. 11 h later the FTOCs were harvested and stained for CD4 and CD8 expression as well as 7AAD to determine DNA content. (A) Dot plots show CD4 and CD8 expression on single cell suspensions of thymocytes from F5/RAG-1^−/− mice–derived FTOCs with or without bcl-2 transgene, with or without addition of the NP68 peptide. Histograms show the DNA content of the DP thymocytes as determined by 7AAD staining. The marker and accompanying figure indicate the percentage of hypodiploid DNA and hence, give a direct measure of the level of apoptosis in each sample. (B) The dot plots and histograms refer to FTOCs from F5/RAG-1^−/− mice with or without the bax transgene.

Figure 4

Bar chart showing the percentage of apoptosis in FTOCs derived from F5/RAG-1^−/− mice plus or minus the bax transgene. The FTOCs were cultured for 4 d before addition of an apoptotic stimulus (+) either 10 μM NP68 peptide, 0.5 μM dexamethasone, or 1 μg/ml anti-Fas antibody (Jo2). After 11 h further culture the FTOCs were harvested and stained with CD4, CD8, and 7AAD. The percentage of apoptosis as determined by hypodiploid DNA content is shown for sets of individual thymic lobes with (+) or without (−) apoptotic stimulus. Each bar represents data from four to six thymic lobes and the error bar represents their standard deviation. The data shown is for a single complete experiment with similar or identical results to four other independent experiments. (Open bars) F5/RAG-1^−/−; (striped bars) F5/RAG-1^−/−/bax.

Figure 5

Bar chart showing the percentage of apoptosis in FTOCs derived from F5/RAG-1^−/− mice plus or minus the bcl-2 transgene. The FTOCs were cultured for 4 d before addition of an apoptotic stimulus (+) either 10 μM NP68 peptide, 0.5 μM dexamethasone, or 1 μg/ml anti-Fas antibody (Jo2). After 11 h further culture the FTOCs were harvested and stained with CD4, CD8, and 7AAD. The percentage apoptosis as determined by hypodiploid DNA content is shown for sets of individual thymic lobes with (+) or without (−) apoptotic stimulus. Each bar represents data from four to six thymic lobes. The data shown is for a single complete experiment with similar or identical results to four other independent experiments. (Open bars) F5/RAG-1^−/−; (striped bars) F5/ RAG-1^−/−/bcl-2.

Figure 6

A model representing the properties of Bax and Bcl-2 in T cell selection. Bax and Bcl-2 have antagonistic effects on positive selection and death by neglect. However, Bax has no effect on negative selection, whereas Bcl-2 can inhibit it. This implies that an unidentified member of the Bcl-2 family can antagonize the effect of Bcl-2 on negative selection.