Reduced Expression of Bcl-2 in CD8⁺ T Cells Deficient in the IL-15 Receptor α-Chain

Generation of IL-15Rα knockout mice

An IL-15Rα genomic clone was isolated from a bacterial artificial chromosome library of RW4 embryonic stem (ES) cells (Genome Systems, St. Louis, MO) derived from 129SvJ mice. A partial map of the IL-15Rα genomic clone was established by restriction endonuclease digestion in combination with Southern blot analysis with a murine IL-15Rα cDNA probe (kindly provided by Immunex, Seattle, WA). An 8.8-kb BamHI-EcoRI genomic fragment that contains exon 1 of the IL-15Rα gene was used to generate the targeting construct. In the targeting construct, a 1.3-kb SmaI-XbaI fragment containing exon 1 of the IL-15Rα gene was replaced by a 2-kb fragment containing the neomycin phosphotransferase gene (neo) under the control of the phosphoglyceraldehyde kinase gene (pgk) promoter and flanked by loxP sites. A pgk promoter-driven thymidine kinase gene (tk) cassette was also ligated to the 3′ end of the short homology arm. The targeting construct was introduced into 129SvJ-derived GSI-1 ES cells (Genome Systems) by electroporation, and the cells were then subjected to selection with ganciclovir and G418. The resistant clones were screened by Southern blot hybridization with a probe that flanks the short homology arm of the targeting construct. Two ES clones were identified and used for microinjection of blastocysts. The resultant male chimeric mice were bred with C57BL/6 (B6) females, and offspring that exhibited germline transmission of the mutant allele (IL-15Rα^+/− mice) were bred with EIIa-cre transgenic mice (18) to delete the neo cassette. The IL-15Rα^+/− neo^−/− offspring were identified by Southern blot analysis and bred with B6 mice. The resultant IL-15Rα^+/− offspring were intercrossed to generate IL-15Rα^−/− and IL-15Rα^+/+ littermates for experiments. Mice used in this study were 6–8 wk old and had been backcrossed to B6 mice for two to five generations.

Cell preparation

Cells of the thymus, spleen, and LNs were released gently from each organ with the use of a glass tissue grinder. RBCs in the spleen cell preparation were lysed by incubation in ACK buffer (0.15 M NH₄Cl, 1 mM KHCO₃, and 0.1 mM EDTA (disodium salt), pH 7.2). Large pieces of tissue debris were removed from the single-cell suspension by gravity sedimentation. PBMCs were prepared from tail blood after lysis of RBCs with ACK buffer. For purification of CD44^lowCD8⁺ cells, LN and spleen cells were first enriched in CD8⁺ cells by negative panning on plates coated with anti-MHC class II mAb (clone BP107; prepared in our laboratory) and then on plates coated with anti-CD4 mAb (clone GK1.5; prepared in our laboratory). The nonadherent cells were collected, stained with a PE-conjugated anti-CD8α (clone 53-6.7; BD Biosciences, San Diego, CA) and an FITC-conjugated anti-CD44 (clone CT-CD4; CALTAG Laboratories, Burlingame, CA) mAb, and sorted for CD44^lowCD8⁺ cells with a FACStar^Plus (BD Biosciences).

Immunostaining and flow cytometry

Staining of cell surface molecules with mAbs was performed for 15 min at room temperature in staining buffer: Mg²⁺- and Ca²⁺-free Dulbecco’s PBS (Life Technologies, Rockville, MD) containing 1% heat-inactivated FCS and 0.1% NaN₃. For intracellular staining of Bcl-2, cells were first stained with mAbs specific for surface molecules and then fixed with 4% paraformaldehyde (Sigma-Aldrich, St. Louis, MO) at 4°C for 20 min. The fixed cells were washed twice with permeabilization buffer (staining buffer plus 0.1% saponin (Sigma-Aldrich)) and then incubated for 30 min at 4°C with an FITC-conjugated mAb to Bcl-2 (clone 3F11; BD PharMingen, San Diego, CA) in permeabilization buffer; an FITC-conjugated isotype-matched mAb (clone A19-3; BD PharMingen) was used as a negative control for intracellular staining. Three-color staining was performed with mAbs conjugated with FITC, PE, or biotin, followed by incubation with streptavidin-allophycocyanin (CALTAG Laboratories), and the samples were analyzed with a FACSCalibur (BD Biosciences). Four-color staining was performed with mAbs conjugated with FITC, PE, Cy5, or biotin, followed by incubation with streptavidin-Texas Red (Zymed Laboratories, South San Francisco, CA), and the samples were analyzed with a FACStar^Plus (BD Biosciences). Data were acquired with the use of CellQuest and analyzed by FlowJo (Tree Star, San Carlos, CA) softwares. Abs used include biotin-conjugated anti-CD8α (clone CT-CD8α; CALTAG Laboratories), Cy5-conjugated anti-CD8α (clone 53-6.7; BD Biosciences), biotin-conjugated anti-CD4 (clone RM4-4; BD PharMingen), biotin-conjugated anti-CD19 (clone 1D3; BD PharMingen), FITC-conjugated anti-CD19 (clone MB19-1; BD Biosciences), PE-conjugated anti-CD25 (clone PC61.5.3; CALTAG Laboratories), FITC-conjugated anti-CD44 (clone IM7; BD Biosciences), biotin-conjugated anti-CD44 (clone IM7; BD PharMingen), PE-conjugated anti-CD69 (clone H1.2F3; BD Biosciences), biotin-conjugated anti-CD69 (clone H1.2F3; BD Biosciences), biotin-conjugated anti-CD45Rb (clone 23G2; BD PharMingen), PE-conjugated anti-CD62L (clone MEL-14; CALTAG Laboratories), PE-conjugated anti-IL-2Rβ (clone TM-1; BD PharMingen), biotin-conjugated DX5 (BD Biosciences), PE-conjugated anti-TCRβ (clone H57-597; BD Biosciences), and FITC-conjugated TCRβ (clone H57-597; prepared in our laboratory).

In vitro activation of naive CD8⁺ cells

Wells of 96-well plates were coated with anti-TCRβ mAb (0.5 μg/well) (clone H57-597; prepared in our laboratory) alone or in combination with anti-CD28 mAb (1 μg/well) (clone 37.51; prepared in our laboratory). Sorted CD44^lowCD8⁺ cells were activated for the indicated times in the coated wells at a density of 10⁴ cells/well in 200 μl of RPMI 10 (RPMI 1640 (Life Technologies) supplemented with 2 mM l-glutamine, 20 mM HEPES-NaOH (pH 7.2), penicillin-streptomycin (2000 U/L), 50 μM 2-ME, and 10% FCS). Recombinant mouse IL-2 (R&D Systems, Minneapolis, MN) or recombinant human IL-15 (R&D Systems) was included in the culture as indicated.

Apoptosis assay

Cells were incubated for 20 min at room temperature with FITC-conjugated annexin V (1 μl/1 × 10⁶ cells; CLONTECH Laboratories, Palo Alto, CA) and propidium iodide (1 μg/ml; Sigma-Aldrich) in annexin V binding buffer (10 mM HEPES-NaOH (pH 7.4), 140 mM NaCl, and 5 mM CaCl₂) and were then analyzed immediately with a FACSCalibur. Annexin V-positive cells were scored as dead cells. Data were acquired with the use of CellQuest and analyzed by FlowJo softwares.

Statistical analysis

Data are expressed as means or means ± SD and were compared among groups by single-classification ANOVA. A value of p < 0.05 was considered to be statistically significant.

Generation of IL-15Rα^−/−mice

The targeting construct for generation of IL-15Rα knockout mice was designed to replace a 1.3-kb region of the IL-15Rα gene containing exon 1 with a 2-kb loxP-flanked neo cassette by homologous recombination (Fig. 1⇓A). Such recombination should give rise to a null allele, given that deletion of exon 1 removes both the translational start codon and the nucleotide sequence encoding the signal peptide. The neo cassette was deleted by breeding IL-15Rα^+/− neo^+/− mice with EIIa-cre transgenic mice. Mouse genotypes were determined by Southern blot hybridization of genomic DNA with a 0.94-kb probe that flanks the 3′ end of the short homology arm of the targeting construct (Fig. 1⇓A). The probe hybridizing with 6.5-, 7.2-, and 5.2-kb BamHI fragments corresponds to the wild-type, IL-15Rα⁻neo⁺, and IL-15Rα⁻neo⁻ alleles, respectively (Fig. 1⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

Generation of IL-15Rα knockout mice. A, Targeting strategy for disruption of the IL-15Rα gene. The targeting construct was designed to replace a 1.3-kb SmaI-XbaI fragment containing exon 1 of the IL-15Rα gene with a 2-kb loxP-flanked pgk-neo cassette by homologous recombination. A pgk-tk cassette was inserted at the 3′ end of the short homology arm of the targeting construct. Restriction enzyme sites: B, BamHI; Sm, SmaI; Xb, XbaI; E, EcoRI. B, Analysis of mutant mouse genotype. Genomic DNA prepared from mouse tail was digested with BamHI and subjected to Southern blot analysis with a 0.94-kb probe (shown in A) that flanks the short homology arm of the targeting construct. Wild-type (WT) and knockout (KO) alleles of the IL-15Rα gene generate 6.5- and 7.2-kb hybridizing fragments, respectively (left panel). The neo cassette of the targeted allele was deleted by breeding the mutant mice with EIIa-cre transgenic mice. The IL-15Rα⁻neo⁻ allele gives rise to a 5.2-kb BamHI fragment that hybridizes with the same 0.94-kb probe (right panel). Symbols above the lanes of both blots indicate the IL-15Rα genotype.

Peripheral lymphocyte profile

The lymphocyte profile of secondary lymphoid organs was analyzed to determine whether IL-15Rα deficiency affects lymphocyte development or maintenance. IL-15Rα^−/− mice exhibited similar numbers of splenic white cells but 30% fewer LN cells (p < 0.1) when compared with wild-type littermates (our unpublished data). The percentage of B cells (CD19⁺) was normal in the knockout mice, whereas the percentages of NK T cells (CD3⁺DX5⁺) and NK cells (CD3⁻DX5⁺) were greatly decreased (our unpublished data). Within the T cell compartment of IL-15Rα^−/− mice, the percentage of CD8⁺ cells was reduced by 50% whereas that of CD4⁺ cells was increased by 20%, resulting in similar percentages of CD4⁺ plus CD8⁺ cells in wild-type and knockout animals (Fig. 2⇓A). Among CD8⁺ T cells, a 72% decrease in the CD44^high population, which is enriched in memory cells, and a 40% decrease in the CD44^low population, which is enriched in naive cells, were detected in the knockout mice (Fig. 2⇓B). The marked deficiency of NK, NKT, and CD44^highCD8⁺ cells apparent in our IL-15Rα^−/− mice is consistent with the characteristics of the IL-15Rα and IL-15 knockout mice described previously (16, 17). Most, if not all, memory CD8⁺ cells generated in vivo are thought to express high levels of both CD44 and IL-2Rβ (19, 20). Moreover, the IL-15-induced proliferation of CD44^highCD8⁺ cells in vivo is largely confined to the IL-2Rβ^high subpopulation (21). Indeed, IL-2Rβ^high cells constituted nearly all CD44^highCD8⁺ cells in wild-type mice and were preferentially reduced in IL-15Rα^−/− mice (Fig. 2⇓C). Blood lymphocyte profile in IL-15Rα^−/− mice was also compared with that in wild-type mice. The percentages of total and CD44^highIL-2Rβ^highCD8⁺ T cells (Fig. 3⇓) as well as those of NK T and NK cells (our unpublished data) were reduced in the blood to similar extents as those apparent in the spleen and LNs of the mutant mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Deficiency of CD8⁺ cells in the secondary lymphoid organs of IL-15Rα^−/− mice. A, Percentages of CD4⁺, CD8⁺, and CD4⁺ plus CD8⁺ cells among total spleen or LN cells of IL-15Rα^−/− mice (○) and wild-type littermates (•) as determined by surface staining with specific mAbs and flow cytometry. Each symbol corresponds to a different animal, and mean values are indicated by horizontal lines. ∗, p < 0.01; ∗∗, p < 0.001 (single-classification ANOVA). B, Flow cytometric analysis of the percentages of CD44^high and CD44^lowCD8⁺ cells among total LN cells of IL-15Rα knockout (−/−) and wild-type (+/+) mice. Data shown are representative of 15 independent experiments. C, Percentages of CD44^highIL-2Rβ^high and CD44^highIL-2Rβ^low cells among CD8⁺ LN cells. Similar results were obtained in five independent experiments.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

Deficiency of CD8⁺ cells in the blood of IL-15Rα^−/− mice. A, Percentages of CD4⁺, CD8⁺, and CD4⁺ plus CD8⁺ cells among PBMCs of IL-15Rα^−/− mice (○) and wild-type littermates (•). Mean values are indicated by horizontal lines. ∗, p < 0.01; ∗∗, p < 0.001 (single-classification ANOVA). B, Percentages of CD44^high and CD44^lowCD8⁺ cells among PBMCs of IL-15Rα knockout (−/−) and wild-type (+/+) mice. Data are representative of four independent experiments. C, Percentages of CD44^highIL-2Rβ^high cells among CD8⁺ cells. Similar results were obtained in four independent experiments.

Thymocyte profile

Given that the percentage of naive CD8⁺ T cells in the periphery of IL-15Rα^−/− mice was reduced by 40%, we next analyzed thymocyte populations to determine whether this deficiency resulted during T cell development in the thymus. IL-15Rα^−/− mice exhibited normal numbers of thymocytes with a normal composition of CD4 and CD8 double-negative (DN) (6), double-positive (DP) (6), and SP populations (our unpublished data). The composition of subpopulations of DN thymocytes defined by expression of CD44 and CD25 also appeared to be normal in the IL-15Rα^−/− mice (our unpublished data). The TCR^high phenotype marks thymocytes that have undergone positive selection (22). Given that ∼35% of CD8SP thymocytes normally exhibit no or only a low level of TCR expression, we examined the percentage of CD8SP TCR^high cells among SP TCR^high thymocytes and found it reduced by ∼20% in IL-15Rα^−/− mice compared with that in wild-type mice (Fig. 4⇓A). CD69 expression marks thymocytes that are undergoing or have just undergone positive selection (23). To examine whether the decrease in the percentage of CD8SP TCR^high thymocytes in IL-15Rα^−/− mice is associated with the DP to SP transition, we examined CD69 expression in TCR^med-high DP and in TCR^high CD8SP thymocytes. The percentages of CD69⁺ cells in these two thymocyte populations of IL-15Rα^−/− mice were similar to the corresponding values for wild-type mice (Fig. 4⇓B). These results suggest that a normal number of DP thymocytes undergoes positive selection and progresses to the CD8SP stage in IL-15Rα^−/− mice, and that the decrease in the percentage of CD8SP TCR^high thymocytes in the knockout animals occurs after the DP to SP transition.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

Deficiency of CD8SP TCR^high thymocytes in IL-15Rα^−/− mice. A, Reduction in the percentage of CD8SP TCR^high thymocyte in IL-15Rα^−/− mice. Total thymocytes from IL-15Rα^−/− mice (○) and wild-type littermates (•) were stained with mAbs specific for CD4, CD8, and TCRβ. The percentages of total CD8SP TCR^high thymocytes among total SP TCR^high thymocytes were indicated. Mean values are provided and indicated by horizontal lines. ∗, p < 0.05 (single-classification ANOVA). B, Similar percentages of CD69⁺ cells among DP or CD8SP TCR^high thymocytes of IL-15Rα^−/− mice and of wild-type mice. Total thymocytes of IL-15Rα knockout (−/−) and wild-type (+/+) mice were stained with mAbs specific for CD4, CD8, TCRβ, and CD69. DP and CD8SP thymocytes were gated for analysis of CD69 and TCRβ expression. Data are representative of five independent experiments.

Reduced abundance of Bcl-2 in CD8⁺ T cells of IL-15Rα^−/− mice

The marked apoptosis and reduction in the number of thymocytes apparent in Bcl-2^−/− mice demonstrate the importance of Bcl-2 for thymocyte survival (24). The level of Bcl-2 in normal thymocytes is tightly regulated (25, 26). The abundance of this anti-apoptotic protein in DN thymocytes is greater than that in DP TCR^low thymocytes at a stage before selection. Down-regulation of Bcl-2 in DP TCR^low thymocytes is thought to facilitate positive selection by promoting the death of cells that fail this process. Thymocytes that have successfully undergone positive selection up-regulate both Bcl-2 and TCR expression and differentiate into TCR^high DP and then TCR^high SP cells (22). To determine whether the deficiency of CD8SP TCR^high thymocytes in IL-15Rα^−/− mice is related to defective cell survival, we examined the abundance of Bcl-2 in SP thymocytes and the preceding DP cells. The amount of Bcl-2 was normal in TCR^low and TCR^high DP thymocytes as well as in CD4SP thymocytes of IL-15Rα^−/− mice, but it was reduced in CD8SP TCR^high thymocytes of these animals (Fig. 5⇓). These results demonstrate that IL-15Rα deficiency results in a reduced concentration of Bcl-2 specifically in CD8SP TCR^high thymocytes that are destined soon to be exported to the periphery to join the naive CD8⁺ T cell pool. The abundance of Bcl-2 in peripheral naive CD8⁺ T cells of IL-15Rα^−/− mice was also reduced compared with that in the corresponding wild-type cells (Fig. 6⇓A), indicating that the Bcl-2^low phenotype is acquired during the development of CD8⁺ cells in the thymus and is conserved in the periphery. Similar to the situation with thymocytes, no difference in Bcl-2 abundance in naive CD4⁺ cells was detected between IL-15Rα^−/− and wild-type mice (Fig. 6⇓A).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

Reduced expression of Bcl-2 in CD8SP TCR^high thymocytes of IL-15Rα^−/− mice. Thymocytes were examined for surface CD4, CD8, and TCRβ as well as intracellular Bcl-2 expression by staining with specific mAbs and analysis by flow cytometry. Filled and open peaks demarcated by solid lines represent Bcl-2 staining of wild-type and IL-15Rα^−/− cells, respectively. Peaks demarcated by dashed lines represent background fluorescence generated by an isotype-matched control mAb in wild-type (thick lines) or IL-15Rα^−/− (thin lines) cells. Only one background peak is shown when wild-type and IL-15Rα^−/− cells exhibited identical background fluorescence. Data are representative of results obtained with three different pairs of mice.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

Reduced expression of Bcl-2 in peripheral CD8⁺ T cells of IL-15Rα^−/− mice. Total LN and spleen cells were stained with mAbs specific for CD8, CD44, and Bcl-2 or specific for CD4, CD45Rb, and Bcl-2, and were then analyzed by flow cytometry. A, Bcl-2 expression by naive (CD44^low) CD8⁺ cells and by naive (CD45Rb^high) CD4⁺ cells. Open and filled peaks demarcated by solid lines represent IL-15Rα^−/− and wild-type cells, respectively. Peaks demarcated by dashed lines represent the background fluorescence generated by an isotype-matched control mAb; given that IL-15Rα^−/− and wild-type cells yielded identical background fluorescence, only one peak is shown. B, Bcl-2 expression by naive (CD44^low) vs memory (CD44^high) CD8⁺ T cells of wild-type (+/+) and IL-15Rα^−/− mice. Open and filled peaks demarcated by solid lines represent naive and memory populations, respectively. Peaks demarcated by dashed lines represent background fluorescence generated by an isotype-matched control mAb. Data in A and B are representative of five independent experiments.

The expression of Bcl-2 was shown to increase in Ag-specific memory CD8⁺ cells after an immune response in vivo, and this effect was suggested to contribute to the long-term maintenance of memory CD8⁺ cells (27). Although the percentages of CD44^highIL-2Rβ^highCD8⁺ cells were greatly reduced, small numbers of these cells were detected in IL-15Rα^−/− mice. We therefore examined whether Bcl-2 expression was up-regulated in these residual memory-like CD8⁺ cells. The abundance of Bcl-2 in CD44^highCD8⁺ cells of IL-15Rα^−/− mice was as low as that in naive CD8⁺ cells of these animals, whereas Bcl-2 expression in CD44^highCD8⁺ cells of wild-type mice was markedly up-regulated compared with that in the corresponding naive CD8⁺ cells (Fig. 6⇑B). These observations thus indicate that the up-regulation of Bcl-2 expression in memory CD8⁺ T cells is dependent on IL-15Rα.

Increased sensitivity of CD8⁺ cells from IL-15Rα^−/− mice to death induction

Bcl-2 protects T cells from apoptosis induced by various stimuli (28, 29). Given that the abundance of Bcl-2 in CD8⁺ thymocytes and peripheral T cells of IL-15Rα^−/− mice is reduced compared with that in the corresponding wild-type cells, we examined the sensitivity of these cells to the induction of apoptosis. Thymocytes and resting T cells undergo apoptosis when removed from the stromal environment in vivo and cultured in vitro. This so-called “spontaneous cell death” is likely due to a lack of survival signals provided either by soluble factors present in the stromal environment or by direct contact between T cells and stromal cells or the extracellular matrix. Thymocytes are also highly sensitive to glucocorticoid-induced cell death. We therefore compared the death susceptibilities of IL-15Rα^−/− and wild-type thymocytes by culturing the cells in medium alone or in medium containing various concentrations of dexamethasone (Fig. 7⇓). In the absence or presence of dexamethasone, CD8SP thymocytes from IL-15Rα^−/− mice exhibited a greater incidence of cell death than did their wild-type counterparts, whereas death rates were similar for either CD4SP or DP thymocytes from the knockout and wild-type mice. Thus, CD8SP, but not CD4SP or DP, thymocytes from IL-15Rα^−/− mice were more susceptible to both spontaneous and dexamethasone-induced death than were the corresponding wild-type cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

Increased sensitivity of CD8SP thymocytes of IL-15Rα^−/− mice to induction of cell death. Total thymocytes isolated from wild-type (+/+) and IL-15Rα knockout (−/−) mice were cultured for the indicated times in RPMI 10 (medium) in the absence or presence of 1 or 10 nM dexamethasone (dex). The percentage of apoptotic cells among CD4SP, CD8SP, and DP populations was determined by staining with mAbs to CD4 and to CD8 in combination with annexin V and propidium iodide. Annexin V-positive cells were scored as dead cells. Data are means ± SD of triplicate determinations. Similar results were obtained in three independent experiments.

We also examined the sensitivity of CD8⁺ LN cells fromIL-15Rα^−/− mice to death induction. The percentages of dead cells among CD44^lowCD8⁺ populations of freshly isolated LN cells were similar for wild-type and IL-15Rα^−/− mice (Fig. 8⇓A). In contrast, the percentage of dead cells among CD44^highCD8⁺ cells was greater for IL-15Rα^−/− mice than for wild-type mice, indicative of poor survival of IL-15Rα memory CD8⁺ cells in vivo. Culture of LN cells in medium alone revealed that the percentage of dead cells among the CD44^highCD8⁺ population of wild-type mice was still smaller than that for their IL-15Rα^−/− counterparts at 24 h but was similar to the IL-15Rα knockout value at 48 h (Fig. 8⇓A). In contrast, the percentages of dead cells among CD44^lowCD8⁺ populations were similar for wild-type and mutant mice at all time points. We then examined TCR-mediated cell death in CD44^lowCD8⁺ cells by stimulation either with immobilized anti-TCRβ and anti-CD28 mAbs or with the immobilized anti-TCRβ mAb plus IL-2 (Fig. 8⇓B). Each treatment resulted in a higher percentage of dead cells for IL-15Rα^−/− mice than for wild-type mice. These results thus indicate that CD44^lowCD8⁺ cells of IL-15Rα^−/− mice survive normally in the resting state despite their reduced Bcl-2 content, but that these cells are more sensitive to TCR-mediated death than are the corresponding wild-type cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 8.

Increased sensitivity of CD8⁺ LN cells of IL-15Rα^−/− mice to death induction. A, Spontaneous cell death. Total LN cells from wild-type (+/+) and IL-15Rα knockout (−/−) mice were cultured in RPMI 10 for the indicated times. The percentage of dead cells in CD44^low and CD44^highCD8⁺ populations was determined by staining with mAbs to CD44 and to CD8 in combination with propidium iodide and annexin V. Data are means ± SD of duplicate determinations, and similar results were observed in two independent experiments. B, TCR-mediated death. CD44^lowCD8⁺ cells were purified by sorting and stimulated for the indicated times with an immobilized mAb to TCRβ in the presence either of an immobilized mAb to CD28 or of IL-2. Dead cells were detected by staining with propidium iodide and annexin V. Data are means ± SD of triplicate determinations, and similar results were observed in four independent experiments.

Up-regulation of Bcl-2 expression in CD8⁺ cells by exogenous IL-15

The reduced abundance of Bcl-2 in CD8⁺ cells of IL-15Rα^−/− mice suggested that IL-15 regulates Bcl-2 expression in CD8⁺ T cells. To investigate this possibility further, we monitored both Bcl-2 concentration and cell death in naive CD8⁺ cells stimulated with a mAb to TCRβ in the presence of various concentrations of exogenous IL-15 (Fig. 9⇓). IL-15 both increased the expression of Bcl-2 and reduced the incidence of cell death among activated wild-type or IL-15Rα CD8⁺ cells in a concentration-dependent manner. The up-regulation of Bcl-2 expression to similar levels required higher concentrations of IL-15 for IL-15Rα^−/−cells than for wild-type cells, consistent with that IL-15Rα or IL-15Rαβγ_c binds IL-15 with high affinity, whereas IL-15R βγ_c binds IL-15 with intermediate affinity (2, 3). These results thus demonstrated a direct effect of IL-15 on Bcl-2 expression in activated CD8⁺ T cells as well as a correlation between Bcl-2 expression and the survival of activated CD8⁺ T cells.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 9.

Effects of exogenous IL-15 on Bcl-2 expression in and survival of CD8⁺ T cells. Sorted CD44^lowCD8⁺ LN cells from wild-type (+/+) and IL-15Rα knockout (−/−) mice were stimulated with immobilized anti-TCRβ mAb in the presence of various concentrations of IL-15 for 48 h. Bcl-2 expression was then determined by intracellular staining with a mAb to this protein; the mean fluorescence intensity (MFI) of staining shown has been corrected for the mean fluorescence intensity obtained with an isotype-matched control mAb. Dead cells were detected by staining with propidium iodide and annexin V, and the data are means of duplicates. Similar results were observed in three independent experiments.