The precursors of TCRαβ+CD8αα+ intraepithelial lymphocytes (IEL) arise in the thymus through a complex process of agonist selection. We and others have shown that the proapoptotic protein, Bim, is critical to limit the number of thymic IEL precursors (IELp), as loss of Bim at the CD4+CD8+ double-positive stage of development drastically increases IELp. The factors determining this cell death versus survival decision remain largely unknown. In this study, we used CD4CreBcl2f/f mice to define the role of the antiapoptotic protein Bcl-2 and CD4CreBcl2f/fBimf/f mice to determine the role of Bcl-2 in opposing Bim to promote survival of IELp. First, in wild-type mice, we defined distinct subpopulations within PD-1+CD122+ IELp, based on their expression of Runx3 and α4β7. Coexpression of α4β7 and Runx3 marked IELp that were most dependent upon Bcl-2 for survival. Importantly, the additional loss of Bim restored Runx3+α4β7+ IELp, showing that Bcl-2 antagonizes Bim to enable IELp survival. Further, the loss of thymic IELp in CD4CreBcl2f/f mice also led to a dramatic loss of IEL in the gut, and the additional loss of Bim restored gut IEL. The loss of gut IEL was due to both reduced seeding by IELp from the thymus as well as a requirement for Bcl-2 for peripheral IEL survival. Together, these findings highlight subset-specific and temporal roles for Bcl-2 in driving the survival of TCRαβ+CD8αα+ IEL and thymic IELp.

This article is featured in Top Reads, p.

Intraepithelial lymphocytes (IEL) are a heterogeneous population of immune cells important for regulating intestinal homeostasis (14). Recent studies also revealed the impact of IEL on metabolism and cardiovascular health (5). IEL are broadly classified into natural and induced IEL. Induced IEL comprise TCRαβ+CD4+ and TCRαβ+CD8αβ+ IEL that arise from peripheral T cells. Natural IEL include TCRαβ+ and TCRγδ+ IEL that do not express the classical coreceptors, but instead express the CD8αα homodimer and arise from precursor cells that differentiate in the thymus (68).

During development, CD4 and CD8 single-positive thymocytes go through the processes of positive and negative selection that generate CD4+ and CD8+ T cells with diverse TCR repertoires. Other cells, including NKT cells, regulatory T cells (Treg), and precursors of TCRαβ+CD8αα+ IEL undergo an ill-defined process of thymic agonist selection, an alternative cell fate associated with strong TCR signals (912). Recent data show that the repertoire of IEL precursors (IELp) is controlled after β-selection and before classical positive selection (13), likely by pre-TCR signal strength (14). In addition to this early repertoire selection step, most IELp also progress through a CD4+CD8+ double-positive (DP) stage as shown by fate mapping mice (15). Developmentally, a subset of DP cells expressing CD8αα (CD4+CD8αβ+CD8αα+ triple-positive [TP] cells) and very little TCRβ and CD5 are the earliest identified preselection IELp to date (16). IELp are thought to further progress when DP cells recognize self-ligands with high-affinity TCR interactions, downregulate CD4 and CD8, and upregulate markers of high TCR signaling, like programmed cell death 1 (PD-1), to become part of the postselected (TCRβ+CD5+), CD4CD8 double-negative (DN) thymocyte pool (9, 1620). Within these postselected DN cells, those expressing CD122, a marker of self-reactive, mature cells (21, 22), have the strongest propensity to populate the small intestine and become TCRαβ+CD8αα+ IEL (22, 23). Recent studies examining IELp heterogeneity identified CD122+ subsets based on their expression of PD-1 and Tbet that vary in MHC restriction, TCR usage, emigration, integrin expression (2225), and timing of gut seeding (26). Although progress has been made in characterizing thymic IELp subsets, the factors that control their survival are just beginning to be elucidated.

Apoptosis plays a crucial role in cell fate and survival decisions in the thymus and is initiated by the interpretation of TCR signals by pro- and antiapoptotic proteins of the Bcl-2 family that physically interact to regulate cytochrome c release from the mitochondria (27, 28). Bim plays a critical role in thymic development, controlling negative selection (29) as well as IELp development (20, 22, 30). Although we previously showed that Bcl-2 plays a critical role in opposing Bim to promote memory CD8+ T cell survival (3133), the factors that antagonize Bim to promote IELp survival remain unclear. Indeed, despite the normal presence of Bim, it remains unclear how some IELp survive. Although transgenic overexpression of prosurvival Bcl-2 family proteins resulted in increased thymic IELp and TCRαβ+CD8αα+ IEL, similar to the loss of Bim, these studies did not examine the natural antagonists of Bim that control IELp survival (20, 3436). Thus, the role of Bcl-2 as a critical Bim antagonist in thymic IELp remains to be rigorously tested.

To avoid substantial issues in global Bcl-2–deficient mice (37), we used CD4CreBcl2f/f mice, in which Bcl-2 was selectively deleted at the DN to DP transition, and found a marked reduction in a subset of postselected DN thymocytes that express PD-1 and CD122. Further, within PD-1+CD122+ thymocytes, we identified a population that had recently experienced TCR signals and a more mature population, expressing Runx3 and enriched for markers of IELp, which critically depended on Bcl-2 to combat Bim for survival. Bcl-2–mediated IELp survival was independent of IL-2 and IL-15 cytokine signals. Importantly, in addition to its role in thymic IELp survival, we also found that Bcl-2 was critical for peripheral survival of gut IEL, and both roles contributed to an almost complete lack of TCRαβ+CD8αα+ IEL in the gut. Thus, although the loss of Bcl-2 was imperceptible at the whole DN population, a small subpopulation of IELp thymocytes was drastically reduced, which contributed to the acute loss of TCRαβ+CD8αα+ IEL in CD4CreBcl2f/f mice.

C57BL/6 mice were purchased from Taconic Farms. Bcl2f/f mice were a gift from Dr. Ira Tabas (Columbia University, New York, NY) and were crossed to CD4Cre+ mice obtained from The Jackson Laboratory. Bimf/f mice were generated in collaboration with Dr. Philippe Bouillet (38, 39) and crossed to CD4Cre+ mice obtained from The Jackson Laboratory. Nur77GFP mice (C57BL/6-Tg [Nr4a1-EGFP/cre] 820Khog/J) were obtained from The Jackson Laboratory. CD4CreBcl2f/f mice were crossed to CD4CreBimf/f mice. Mice for the thymus studies were between 3 and 5 wk of age and for the gut studies between 5 and 8 wk of age. Experimental and control mice were littermates in most cases, and both male and female mice were used. Wild-type (WT) mice mentioned in the figure legends are C57BL/6, Bcl2f/f, Bimf/f, or Bcl2f/f Bimf/f mice; we have not observed significant differences in any of these lines of mice. Animals were housed under specific pathogen-free conditions in the Division of Veterinary Services (Cincinnati Children’s Hospital Medical Center, Cincinnati, OH), and experimental procedures were reviewed and approved by the Institutional Animal Care and Use Committee at the Cincinnati Children’s Hospital Research Foundation.

Thymi or spleens from individual mice were harvested and crushed through a 100-μm mesh strainer and RBCs were lysed to generate single-cell suspensions. IEL isolation was performed by following a modified protocol (40). Briefly, the small intestine was cut off and washed with Ca2+Mg2+-free HBSS for removing feces and mucus. After removing Peyer’s patches, the small intestine was cut in pieces and shaken in Ca2+Mg2+-free HBSS supplemented with 5% FBS, 2 mM EDTA, and 1 mM DTT (Sigma-Aldrich), twice, for 15 min at 37°C. The IEL in the supernatant were purified with Percoll (GE Healthcare) gradient. The single-cell suspension was further analyzed by flow cytometry after excluding dead cells by using the Live Dead staining kit (Invitrogen). Two to 3 million cells were stained with tetramers that recognize CD1d (PBS-57) or thymic leukemia Ag (T3b) (National Institutes of Health Tetramer Core Facility) followed by staining with Abs against CD4, TCRβ, CD5, α4β7, Runx3 (BD Biosciences), CD25, CD8α, PD-1, CD122, CD44, CD45.2, CD8β, Bcl-2, Ki-67, TCRδ, CD49b, NKp46 (BioLegend), Bim (Cell Signaling Technology), or Tbet. Intracellular stains were performed using the Foxp3/transcription factor staining buffer set (Invitrogen). For detection of Bim, secondary anti-rabbit IgG Ab was used (Invitrogen). The cells were acquired on a BD LSRFortessa flow cytometer and analyzed by FACSDiva software (BD Biosciences) or FlowJo software (Tree Star).

Anti–IL-2 Ab (clones S4B6 and JES6-1A12) and rat IgG2a isotype control (2A3) were purchased from BioXCell. Anti–IL-15 (M96) was a kind gift from Amgen. For IL-2 and IL-15 neutralization, 3-wk-old mice were injected i.p with 150 μg S4B6 and Jes61A12 and 25 μg M96 or with 150 μg isotype control (2A3) on days 0, 2, 4, and 6. The mice were sacrificed on day 7, and thymii and spleen were harvested.

GraphPad Prism and Microsoft Excel software were used to analyze data and generate graphs. Statistical tests were performed as described in the figure legends.

Given the role of Bim in limiting IELp survival (20, 30), we examined the role of a major Bim antagonist, Bcl-2, in promoting IELp survival. To temporally delete Bcl-2 during thymic development, we bred Bcl2f/f (41) mice to CD4Cre+ mice. First, overall numbers of preselection TP (CD4+CD8αβ+CD8αα+) IELp were not affected by the loss of Bcl-2 (Fig. 1A). Next, we assessed DN cells (CD4CD8) after excluding NKT cells (CD1d tetramer+) and potential Treg precursors (CD25+) (22) and did not see a significant loss of DN cell numbers in CD4CreBcl2f/f mice (Fig. 1B). In contrast, the frequency and total numbers of postselected DN cells (CD4CD8CD5+TCRβ+) dropped significantly in CD4CreBcl2f/f mice compared with WT mice (Fig. 1B). This reliance on Bcl-2 was associated with higher Bcl-2 expression in CD5+TCRβ+ DN cells compared with DP cells in WT mice (Fig. 1C). Of note, the levels of Bim were significantly lower in CD5+TCRβ+ DN cells in CD4CreBcl2f/f mice compared with WT mice (Fig. 1C). Thus, postselected DN thymocytes normally express higher levels of Bcl-2 that is required for their accumulation.

FIGURE 1.

Bcl-2 contributes to the survival of CD5+TCRβ+ DN thymocytes. (A) Within CD4+CD8+ thymocytes, TP (CD4+CD8αα+CD8αβ+) preselection precursors of TCRαβ+CD8αα+ IEL were identified by gating on thymus leukemia (TL) Ag tetramer-positive cells and CD8β. (B) CD25CD1dtetCD4CD8 DN (top row) and, within the DN, postselected CD5+TCRβ+ thymocytes (bottom row) were determined. Numbers in representative plots show the percentage and bar graphs show the numbers of each population from Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (C) Histograms show the mean fluorescence intensity of Bcl-2 and Bim in CD5+TCRβ+ DN, CD5TCRβ DN, DP, and CD8+ thymocytes from Bcl2f/f and CD4CreBcl2f/f mice. Results are representative of at least three independent experiments with n = 3 or more mice/group and show mean ± SD. **p < 0.01, Student t test.

FIGURE 1.

Bcl-2 contributes to the survival of CD5+TCRβ+ DN thymocytes. (A) Within CD4+CD8+ thymocytes, TP (CD4+CD8αα+CD8αβ+) preselection precursors of TCRαβ+CD8αα+ IEL were identified by gating on thymus leukemia (TL) Ag tetramer-positive cells and CD8β. (B) CD25CD1dtetCD4CD8 DN (top row) and, within the DN, postselected CD5+TCRβ+ thymocytes (bottom row) were determined. Numbers in representative plots show the percentage and bar graphs show the numbers of each population from Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (C) Histograms show the mean fluorescence intensity of Bcl-2 and Bim in CD5+TCRβ+ DN, CD5TCRβ DN, DP, and CD8+ thymocytes from Bcl2f/f and CD4CreBcl2f/f mice. Results are representative of at least three independent experiments with n = 3 or more mice/group and show mean ± SD. **p < 0.01, Student t test.

Close modal

IELp consist of multiple subpopulations, including self-reactive PD-1+ cells and non–self-reactive Tbet+ cells (22, 24). The frequencies and numbers of both PD-1+ and Tbet+ cells were significantly reduced in CD4CreBcl2f/f mice relative to WT mice (Fig. 2A). Recent work showed that CD122 expression is associated with high-affinity Ag encounter and IELp potential (23) and that PD-1 upregulation precedes CD122 upregulation (42). Using these markers as a temporal guide, we found that ∼45% of PD-1+ cells were CD122+, whereas nearly all Tbet+ cells were CD122+ (Fig. 2B). Compared to WT mice, CD4CreBcl2f/f mice displayed a significant drop in frequency and an ∼5-fold loss in the numbers of CD122+ cells in both PD-1+ and Tbet+ populations (Fig. 2B). In contrast, Bcl-2 was not required to maintain normal numbers of CD122 cells in either the PD-1+ or Tbet+ populations. Again, consistent with their dependence upon Bcl-2, in WT mice, we observed a 3- to 4-fold increase in Bcl-2 levels in PD-1+ cells that coexpressed CD122 (Fig. 2C). The PD-1+CD122+ cells that survived in CD4CreBcl2f/f mice had significantly lower levels of Bim in comparison with WT mice (Fig. 2C). In contrast, Tbet+ cells did not have a significant difference in Bim levels between WT and CD4CreBcl2f/f mice (Fig. 2C).

FIGURE 2.

Bcl-2 is critical for the survival of CD122+ thymocytes. (A) Within CD5+TCRβ+ DN thymocytes, PD-1– and Tbet-expressing populations were identified. (B) Within the PD-1+ and Tbet+ cells, CD122-expressing populations were identified. Numbers in representative plots show the percentage and bar graphs show the numbers from Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (C) Histograms show Bcl-2 and Bim expression in PD-1+CD122, PD-1+CD122+, and Tbet+ populations in Bcl2f/f and CD4CreBcl2f/f mice. Results are representative of at least four independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, Student t test.

FIGURE 2.

Bcl-2 is critical for the survival of CD122+ thymocytes. (A) Within CD5+TCRβ+ DN thymocytes, PD-1– and Tbet-expressing populations were identified. (B) Within the PD-1+ and Tbet+ cells, CD122-expressing populations were identified. Numbers in representative plots show the percentage and bar graphs show the numbers from Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (C) Histograms show Bcl-2 and Bim expression in PD-1+CD122, PD-1+CD122+, and Tbet+ populations in Bcl2f/f and CD4CreBcl2f/f mice. Results are representative of at least four independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, Student t test.

Close modal

The integrin α4β7 is upregulated on some PD-1+CD122+ IELp and necessary for their ability to home and be retained in the gut (23, 43). Further, unlike Tbet+ IELp, no lineage-defining transcription factors have been identified for PD-1+ CD122+ IELp. As Runx3 is a critical driver of CD8α reexpression (44) and differentiation of IELp into TCRαβ+CD8αα+ IEL (35), we explored the expression of Runx3 and α4β7 among PD1+CD122+ thymocytes after excluding CD44hi cells (∼10%) (Supplemental Fig. 1A), as PD-1+ IELp are predominantly negative for CD44 expression (22). Interestingly, Runx3 appeared to be coexpressed with α4β7 in a subpopulation of PD-1+CD122+ cells but not PD-1+CD122 cells (Fig. 3A, Supplemental Fig. 1B). Roughly 30% of PD-1+CD122+ cells expressed both Runx3 and α4β7, ∼16% expressed Runx3 but not α4β7, and ∼10% expressed α4β7 but not Runx3, whereas the rest (∼44%) lacked expression of both markers (Fig. 3A). Bcl-2 was required for normal numbers of all four subpopulations with Runx3+α4β7+ cells being maximally impacted (∼3-fold) (Fig. 3A).

FIGURE 3.

Runx3 expression marks a mature IELp subpopulation among PD-1+ CD122+ thymocytes that critically depends on Bcl-2 for survival. (A) Flow cytometric identification of Runx3α4β7, Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ subpopulations within PD-1+CD122+CD44lo postselected DN thymocytes. The bar graphs show the numbers of each population from Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (B) Representative dot plots and bar graphs show the percentage of CD69+ cells among subpopulations of PD-1+CD122+CD44lo thymocytes in WT mice. (C) Bar graphs show the mean fluorescence intensity (MFI) of Bcl-2 and Bim in the subpopulations of PD-1+CD122+CD44lo thymocytes from WT mice in comparison with PD-1+CD122 cells. Results are representative of three independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001, Student t test.

FIGURE 3.

Runx3 expression marks a mature IELp subpopulation among PD-1+ CD122+ thymocytes that critically depends on Bcl-2 for survival. (A) Flow cytometric identification of Runx3α4β7, Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ subpopulations within PD-1+CD122+CD44lo postselected DN thymocytes. The bar graphs show the numbers of each population from Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (B) Representative dot plots and bar graphs show the percentage of CD69+ cells among subpopulations of PD-1+CD122+CD44lo thymocytes in WT mice. (C) Bar graphs show the mean fluorescence intensity (MFI) of Bcl-2 and Bim in the subpopulations of PD-1+CD122+CD44lo thymocytes from WT mice in comparison with PD-1+CD122 cells. Results are representative of three independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ***p < 0.0001, Student t test.

Close modal

To further discern temporal information regarding these postselected PD-1+CD122+ subpopulations, we examined levels of CD69, a transient indicator of recent TCR stimulation. Runx3α4β7 cells were predominantly CD69hi, whereas Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ cells had low expression of CD69 (Fig. 3B). Prior data indicate that CD122+ cells that are α4β7hi and CD69lo are temporally the most distant from the DP stage (22, 23), suggesting that Runx3+α4β7+ cells are the most differentiated IELp subset and are poised to emigrate to the gut. Also, the levels of Bcl-2 were highest in Runx3+α4β7+ cells, followed by Runx3+α4β7 and Runx3α4β7+ cells, whereas Runx3α4β7 cells expressed Bcl-2 at similar levels to PD-1+CD122 cells (Fig. 3C). Runx3+α4β7+ IELp also expressed moderate levels of CD5 and Nur77 [measured in a Nur77GFPreporter mouse (45)] and high CD122 in comparison with the least mature CD122+ IELp that lacked expression of both Runx3 and α4β7 (Supplemental Fig. 2).

As common γ-chain cytokine signaling induces Bcl-2 expression (46), we examined the role of IL-2/15 signaling on Bcl-2 levels and PD-1+CD122+ IELp survival by neutralizing IL-2 and IL-15. The efficiency of neutralization was confirmed by assessing NK cell and Treg depletion in the spleen (Supplemental Fig. 3A, 3B). IL-2/15 neutralization did not impact Bcl-2 levels (Supplemental Fig. 3C), and a small proportion of Runx3+α4β7+ IELp showed reduced proliferation, but their overall numbers were not reduced (Supplemental Fig. 3D). Combined, these data show that Bcl-2 is critical for the survival of CD122+ cells, particularly Runx3+α4β7+ IELp, in an IL-2– and IL-15–independent manner.

One function of Bcl-2 is physical sequestration and antagonization of Bim (28, 47). Interestingly, and similar to Bcl-2, Bim levels appeared highest within Runx3+α4β7+ IELp, slightly lower in Runx3α4β7+ and Runx3+α4β7 cells and even lower in Runx3α4β7 cells (Fig. 3C). Using CD4CreBimf/f mice, we found that the loss of Bim massively increased the accumulation of CD5+TCRβ+ DN and PD-1+ thymocytes (Fig. 4A), but not as much Tbet+ IELp, consistent with prior work (22). Further, loss of Bim led to a significant accrual of CD122+ cells (∼35-fold) and to a lesser extent CD122 cells (∼8-fold) among the PD-1+ population compared with WT (Fig. 4A). We also found that Bim was required to limit the numbers of all four PD-1+CD122+ subpopulations that also required Bcl-2, and the greatest effect was in Runx3α4β7+ cells (Fig. 4B), with ∼290-fold increase compared with WT. Runx3+α4β7+ IELp were also substantially increased (∼180-fold) in the absence of Bim, but despite their increased numbers, they had reduced Runx3 levels (Fig. 4C). Further, these Runx3+α4β7+ cells had very low levels of Bcl-2, suggesting that the absence of Bim allowed visualization of cells that otherwise would have perished (Fig. 4D). Similarly, the loss of Bcl-2 also led to reduced Runx3 levels (Fig. 4C). Taken together, these data show that Bim is required for the clonal deletion of CD122+ IELp encountering strong TCR stimulation and limits their differentiation into Runx3- and α4β7-expressing cells.

FIGURE 4.

Bcl-2 antagonizes Bim to maintain survival of CD122+ IELp. (A) Dot plots show frequencies and bar graphs show thymocyte numbers from WT (open circle), CD4Cre+Bcl2f/f (filled circle), CD4CreBimf/f (open triangle), and CD4CreBimf/f Bcl2f/f (filled square) mice of DN (CD25CD1tetCD4CD8) (row 1), CD5+TCRβ+ among DN (row 2), PD-1+ and Tbet+ among CD5+TCRβ+ DN (row 3), and CD122+ and CD122 among the PD-1+ (row 4). (B) Runx3α4β7, Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ cells within PD-1+CD122+CD44lo thymocytes. (C) Bar graph compares mean fluorescence intensity (MFI) of Runx3 in the Runx3+α4β7+ subpopulation in WT (open circle), CD4Cre+Bcl2f/f (filled circle), and CD4Cre+Bimf/f (open triangle) mice. (D) Bar graphs compare MFI of Bcl-2 in Runx3α4β7, Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ subpopulations between WT (open circle) and CD4CreBimf/f (open triangle) mice and of Bim between WT (open circle) and CD4CreBcl2f/f (filled circle) mice. Results are representative of at least three independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, Student t test.

FIGURE 4.

Bcl-2 antagonizes Bim to maintain survival of CD122+ IELp. (A) Dot plots show frequencies and bar graphs show thymocyte numbers from WT (open circle), CD4Cre+Bcl2f/f (filled circle), CD4CreBimf/f (open triangle), and CD4CreBimf/f Bcl2f/f (filled square) mice of DN (CD25CD1tetCD4CD8) (row 1), CD5+TCRβ+ among DN (row 2), PD-1+ and Tbet+ among CD5+TCRβ+ DN (row 3), and CD122+ and CD122 among the PD-1+ (row 4). (B) Runx3α4β7, Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ cells within PD-1+CD122+CD44lo thymocytes. (C) Bar graph compares mean fluorescence intensity (MFI) of Runx3 in the Runx3+α4β7+ subpopulation in WT (open circle), CD4Cre+Bcl2f/f (filled circle), and CD4Cre+Bimf/f (open triangle) mice. (D) Bar graphs compare MFI of Bcl-2 in Runx3α4β7, Runx3α4β7+, Runx3+α4β7, and Runx3+α4β7+ subpopulations between WT (open circle) and CD4CreBimf/f (open triangle) mice and of Bim between WT (open circle) and CD4CreBcl2f/f (filled circle) mice. Results are representative of at least three independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, Student t test.

Close modal

Consistent with Bcl-2 as a critical antagonist of Bim, the additional loss of Bim largely restored CD5+TCRβ+ DN and PD-1+ cells that lacked Bcl-2 (Fig. 4A). In contrast, Tbet+ cells that were decreased in CD4CreBcl2f/f mice were only partially restored in CD4CreBimf/fBcl2f/f mice. Further, though CD122 and CD122+ cells within the PD-1+ population were increased above WT levels, CD122+ cells were not rescued to the levels seen in CD4CreBimf/f mice (Fig. 4A). Among the CD122+ subpopulations, Runx3α4β7+ cells and Runx3+α4β7+ IELp showed a similar pattern of rescue above WT, but not to the extent observed in the absence of Bim alone (Fig. 4B). Notably, we confirmed that Bim and Bcl-2 were predominantly deleted in all of these populations (Supplemental Fig. 4A). Examination of other Bcl-2 family members in CD122+ cells revealed a difference in the levels of antiapoptotic Mcl-1 and proapoptotic Noxa between CD4CreBimf/f mice and CD4CreBimf/fBcl2f/f mice (Supplemental Fig. 4B). Taken together, our data show that Bcl-2 is critical to counteract the proapoptotic activity of Bim to control the homeostasis of PD-1+CD122+ thymocytes.

Given the requirement of Bcl-2 in thymic development of IELp, we examined whether this translated to TCRαβ+CD8αα+ IEL in the gut. The 8-wk-old CD4CreBcl2f/f mice showed a striking loss (∼30-fold) of TCRαβ+CD8αα+ IEL, whereas TCRαβ+CD8αβ+ IEL showed an ∼5-fold reduction (Fig. 5A). Interestingly, even as early as 5 wk, we found that the few remaining TCRαβ+CD8αα+ IEL in CD4CreBcl2f/f mice had not deleted Bcl-2 despite relatively efficient deletion (∼90%) in thymic IELp as well as in TCRαβ+CD8αβ+ IEL in the gut (Fig. 5B, middle row). As the initial seeding of the gut by thymic IELp occurs between 2-3 wk of age, we examined Bcl-2 expression in IEL in 3-wk-old mice. In contrast to 5-wk-old mice, 3-wk-old CD4CreBcl2f/f mice showed >90% Bcl-2 deletion in TCRαβ+CD8αα+ IEL (Fig 5B, top row). Importantly, the numbers of TCRαβ+CD8αα+ IEL in 3-wk-old CD4CreBcl2f/f mice were significantly decreased, consistent with substantially reduced seeding from the thymus (Fig. 5C). We assessed Ki-67 expression and found no drop in proliferation, excluding impaired proliferation as a mechanism for the reduced numbers (Fig. 5D). Strikingly, we found that the additional loss of Bim completely restored Bcl-2–deleted cells at 5 wk of age (Fig. 5B, bottom row), highlighting the role of Bim in driving the selection of Bcl-2hi cells, and correspondingly, the number of TCRαβ+CD8αα+ IEL were rescued (Fig. 5C). Combined, these data show that both thymic and peripheral Bcl-2 expression critically affects gut IEL survival by combating Bim.

FIGURE 5.

Thymic deletion of Bcl-2 leads to a profound loss of TCRαβ+CD8αα+ IEL. (A) Representative dot plots indicate frequency of gut IEL populations (gated from live, CD45.2+ cells) and show TCRβ+ and TCRδ+ IEL (top row), CD8α and CD4 expression among TCRβ+ IEL (middle row), and CD8β expression among the CD8α+ IEL (bottom row). The bar graph shows the numbers of TCRαβ+CD8αα+, TCRαβ+CD8αβ+, and TCRαβ+CD4+ IEL from 8-wk-old Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (B) Histograms show Bcl-2 expression in TCRαβ+CD8αα+ IEL (left column) and TCRαβ+ CD8αβ+ IEL (right column) from 3-wk-old and 5-wk-old WT (solid line, no fill), CD4CreBcl2f/f (solid line and gray filled) mice and CD4CreBcl2f/f Bimf/f (dashed line and light gray filled) mice. (C) Bar graphs show numbers of IEL in WT (open circle) and CD4CreBcl2f/f (filled circle) at 3 wk of age and CD4CreBimf/f (open triangle) and CD4CreBimf/f Bcl2f/f (filled square) at 5 wk of age. (D) Bar graph shows the frequency of Ki-67+ cells among TCRαβ+CD8αα+, TCRαβ+CD8αβ+, and TCRαβ+CD4+ IEL from 3-wk-old Bcl2f/f mice (open circles) and CD4CreBcl2f/f mice (filled circles). Results are representative of at least three independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, Student t test.

FIGURE 5.

Thymic deletion of Bcl-2 leads to a profound loss of TCRαβ+CD8αα+ IEL. (A) Representative dot plots indicate frequency of gut IEL populations (gated from live, CD45.2+ cells) and show TCRβ+ and TCRδ+ IEL (top row), CD8α and CD4 expression among TCRβ+ IEL (middle row), and CD8β expression among the CD8α+ IEL (bottom row). The bar graph shows the numbers of TCRαβ+CD8αα+, TCRαβ+CD8αβ+, and TCRαβ+CD4+ IEL from 8-wk-old Bcl2f/f (open circle) and CD4CreBcl2f/f (filled circle) mice. (B) Histograms show Bcl-2 expression in TCRαβ+CD8αα+ IEL (left column) and TCRαβ+ CD8αβ+ IEL (right column) from 3-wk-old and 5-wk-old WT (solid line, no fill), CD4CreBcl2f/f (solid line and gray filled) mice and CD4CreBcl2f/f Bimf/f (dashed line and light gray filled) mice. (C) Bar graphs show numbers of IEL in WT (open circle) and CD4CreBcl2f/f (filled circle) at 3 wk of age and CD4CreBimf/f (open triangle) and CD4CreBimf/f Bcl2f/f (filled square) at 5 wk of age. (D) Bar graph shows the frequency of Ki-67+ cells among TCRαβ+CD8αα+, TCRαβ+CD8αβ+, and TCRαβ+CD4+ IEL from 3-wk-old Bcl2f/f mice (open circles) and CD4CreBcl2f/f mice (filled circles). Results are representative of at least three independent experiments with n = 3 or more mice/group and show mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, Student t test.

Close modal

In this study, we identified Bcl-2 as a critical antagonist of Bim in the thymic precursors of TCRαβ+CD8αα+ IEL. The gain and loss of cells in mice deficient in Bcl-2 and/or Bim was instructive in our understanding of thymic CD122+ IELp development. Combined with the expression of maturation markers, our data are consistent with a scenario in which Runx3α4β7 cells are the least mature, PD-1+CD122+ thymocytes having just undergone TCR signaling as they are CD69hi and Nur77hi (45) and minimally affected by the loss of Bcl-2. Runx3α4β7+ and Runx3+α4β7+ cells are more mature as they have lower levels of CD69 and Nur77 and higher levels of α4β7 and are most affected by the loss of Bcl-2. Notably, Runx3+α4β7+ IELp match the phenotype of an agonist-selected cluster recently identified in a thymic single-cell transcriptomic analysis in their expression of Runx3, moderate levels of CD5 and Nur77, high CD122, and low CD69, suggesting that they represent the most mature, agonist-selected, thymic IELp (48).

Although we showed that IL-2/15 signaling was not required for Bcl-2 expression in Runx3+α4β7+ IELp, we think that TCR signals likely contribute to their survival. One scenario for high Bcl-2 acquisition in a subset of CD122+ cells is through pathways specific to agonist signals. For example, mice deficient in RasGRP1 and α-chain connecting peptide domain (α-CPM) have impaired agonist selection (9, 23). Notably, Stim1/Stim2-deficient mice have impaired agonist selection (49) and reduced acquisition of the TCRβhiDPloPD-1hi phenotype (18). The Stim1/Stim2 pathway involving store-operated calcium, triggered upon high-affinity TCR signals, induces transcription factors like Egr2 (49) known to induce Bcl-2 (50). Furthermore, Stim1/Stim2-deficient mice displayed reduced postselected DN numbers and CD122 levels (49). Although the authors attributed this loss to reduced proliferation due to impaired cytokine signaling through CD122, we showed that very few PD-1+CD122+ IELp proliferated in an IL-2/15–dependent manner, and their numbers remain unaffected upon IL-2/IL-15 neutralization in agreement with prior studies (22, 51). Thus, our data are consistent with a model in which TCR signaling primarily upregulates Bcl-2. However, this does not rule out a potential role for Bcl-2 at an earlier stage of IELp development, shortly after β-selection via signals emanating from the pre-TCR (13, 14, 52). Moreover, recent data point to a potential role for TGF-β in maintaining Bcl-2 expression in postselected DN thymocytes (53). Finally, it is also possible that high Bcl-2 levels may be due to escape from B7 signaling as CD28 signaling can reduce Bcl-2 in vitro (54), and B7 knockout mice have impaired clonal deletion of postselected DN (35). It would be interesting to determine whether B7 signals contribute to Bim and Bcl-2 expression in IELp and if Bcl-2 is required for the increased IELp found in B7-deficient animals.

Although Bim kills PD-1+CD122 cells, for which the DPlo and PD-1hi phenotype may be indicative of cells undergoing negative selection (20), its effects were most apparent in PD-1+CD122+ cells, which experience the highest TCR signaling and are associated with IELp agonist selection (23). Prior work has shown that TCR signaling is known to increase Bim levels (29). Our data demonstrate that the only PD-1+CD122+ thymocytes that survive in the presence of Bim are those with high levels of Bcl-2 that enable them to antagonize Bim-mediated death. Indeed, in the absence of Bim, we observed Bcl2lo cells, whereas in the absence of Bcl-2, Bimlo cells emerged. These data show that the levels of Bim and Bcl-2 are tightly aligned and are critical controllers of PD-1+CD122+ IELp survival. However, additional Bcl-2 family members may also play a role in CD122+ IELp homeostasis. For example, the decrease in PD-1+CD122+ cells in CD4CreBcl2f/fBimf/f mice relative to CD4CreBimf/f mice suggests that Bcl-2 antagonizes other proapoptotic proteins in addition to Bim (5557). We examined the levels of additional proapoptotic proteins and found that Noxa levels were significantly lower in PD-1+CD122+ cells in the absence of Bcl-2 (Supplemental Fig. 4B), suggesting Noxa as a potential Bcl-2 antagonist. However, it may be a little more complex than involvement of a single proapoptotic family member, as we found that the levels of antiapoptotic molecule Mcl-1 were also increased in PD-1+CD122+ cells in CD4CreBcl2f/f mice (Supplemental Fig. 4B). We wish to stipulate that these results reflect cells that may have undergone adaptation to the lack of Bim and must be interpreted with that caveat when thinking about roles for additional Bcl-2 family member involvement. Nonetheless, although Bim and Bcl-2 play dominant roles in PD-1+CD122+ IELp, more work is required determine the potential roles of additional Bcl-2 family members and their contribution to thymic IELp homeostasis.

Although our data clearly show that the absence of Bim increases the numbers of PD-1+CD122+ subpopulations that we have defined on the basis of Runx3 and α4β7 expression, the relationship between these subpopulations in the context of IELp differentiation remains unclear. Further work will define the temporal relationships between these subpopulations and their sensitivity to Bim-mediated death. Based on the fact that the levels of Runx3 are lower in Runx3+α4β7+ IELp in the absence of Bim, we hypothesize that the absence of Bim promotes their survival, but may not completely restore differentiation.

Moreover, CD4CreBimf/f mice have a much lower gut increase in TCRαβ+CD8αα+ IEL (∼8-fold) relative to the increase in Runx3+α4β7+ thymic IELp (∼180-fold). One explanation for this could be a reduction in migration and/or survival of IELp as they traffic to the gut. Indeed, on examining the spleen, through which IELp traffic before populating the gut (20, 30), the increase in splenic IELp in CD4CreBimf/f mice was closer to the increase observed in the gut (∼13-fold) (Supplemental Fig. 3E). It is also possible that in the absence of Bim, IEL differentiation is impaired when IELp seed the gut, given that IELp in CD4CreBimf/f mice have significantly lower Runx3 levels and Runx3 is critical for TCRβ+CD8αα+ IEL differentiation (35). Similarly, CD4CreBcl2f/f mice also have low levels of Runx3 in Runx3+α4β7+ IELp, and this likely leads to impaired IEL maturation, which may explain the higher ∼30–40-fold loss of gut TCRβ+CD8αα+ IEL compared with ∼5.5-fold loss of thymic IELp. In further support, transgenic expression of Bcl-2 was unable to rescue IEL deficiency in Runx3-deficient mice (35), consistent with an uncoupling of survival from differentiation. However, the role of Bcl-2 in peripheral survival of TCRαβ+CD8αα+ IEL (evidenced by a temporal selection for Bcl-2hi cells) also contributes to their profound loss in the gut.

Although the importance of Runx3 in the final maturation of IEL in the gut is known (16, 24, 35, 58), our results show that Runx3 expression is actually established in a small subpopulation of CD122+ thymic IELp. This might explain why other groups found that total TCRβ+ DN thymocytes were unchanged in Runx3-deficient mice (35). Similarly, we saw that total numbers of DN cells were not affected in Bcl-2–deficient mice, but a small Runx3+α4β7+ subset was dramatically decreased. Thus, the role of Runx3 in the thymic IELp fate decision needs to be reexamined in light of our work. The signals driving thymic Runx3 expression are yet unknown. Unlike in CD8+ thymocytes, in which common γ-chain cytokines play a role in Runx3 induction, we ruled out IL-2/15–mediated Runx3 upregulation in Runx3+α4β7+ IELp, as their blockade did not affect Runx3 expression (data not shown). Further, IELp do not express the IL-7Rα chain (35), precluding a role for IL-7–mediated Runx3 induction.

Although prior studies used transgenic overexpression of antiapoptotic proteins, our temporally relevant deletion model overcomes the caveats of nonphysiological expression of Bcl-2 family members and potential skewing of interacting partners and reveals Bcl-2 as the critical Bim antagonist involved in both thymic IELp and TCRαβ+CD8αα+ IEL survival. We previously showed that Bim affects the TCR repertoire of postselected DN cells in the periphery (30), and, whereas TCR self-reactivity was conserved in Bcl-2 transgenic mice (34, 59), the effect of loss of Bcl-2 on the TCR repertoire of IELp and IEL requires further study. As prior work associated Bim with inflammatory bowel disease (60), one possible role for Bim (and Bcl-2) is to hone the TCR repertoires of IEL that are critical for regulation of gut injury. As mentioned earlier, repertoire selection may occur prior to classical positive selection (13), and experiments are underway to determine the impact of temporal deletion of Bim and Bcl-2 on thymic IELp repertoires. Nonetheless, as TCRαβ+CD8αα+ IEL have a cytotoxic phenotype, it will be important to test if TCR repertoire changes of IEL impact defense against gut pathogens. Due to the virtual absence of TCRαβ+CD8αα+ IEL in the gut of CD4CreBcl2f/f mice, we think they can be a useful tool to interrogate the role of these cells in vivo, with the caveat that the absence of Bcl-2 modestly reduces conventional CD4+ and CD8+ T cells. Nonetheless, the role of TCRαβ+CD8αα+ IEL in phenotypes uncovered in CD4CreBcl2f/f mice would need to be validated via adoptive transfer of TCRαβ+CD8αα+ IEL from WT mice back into these mice and assessing their contribution to gut homeostasis.

We thank Dr. Ira Tabas (Columbia University, New York, NY) for the kind gift of the Bcl2f/f mice.

This work was supported by U.S. Public Health Service (National Institutes of Health) Grant AI057753, by generous startup funds from the Cincinnati Children’s Hospital Research Foundation, and in part by the Center for Rheumatic Disease Research (National Institutes of Health AR070549).

S.S. designed and performed experiments and analyzed data. S.A.H. performed certain experiments and helped with genotyping of mice. V.V. helped with genotyping of mice. S.S. and D.A.H. wrote the manuscript.

The online version of this article contains supplemental material.

Abbreviations used in this article

DN

double-negative

DP

double-positive

IEL

intraepithelial lymphocyte

IELp

intraepithelial lymphocyte precursor

PD-1

programmed cell death 1

TP

triple-positive

Treg

regulatory T cell

WT

wild-type

1.
Cheroutre
H.
,
F.
Lambolez
,
D.
Mucida
.
2011
.
The light and dark sides of intestinal intraepithelial lymphocytes.
Nat. Rev. Immunol.
11
:
445
456
.
2.
McDonald
B. D.
,
B.
Jabri
,
A.
Bendelac
.
2018
.
Diverse developmental pathways of intestinal intraepithelial lymphocytes.
Nat. Rev. Immunol.
18
:
514
525
.
3.
Hoytema van Konijnenburg
D. P.
,
B. S.
Reis
,
V. A.
Pedicord
,
J.
Farache
,
G. D.
Victora
,
D.
Mucida
.
2017
.
Intestinal epithelial and intraepithelial T cell crosstalk mediates a dynamic response to infection.
Cell
171
:
783
794.e13
.
4.
Sujino
T.
,
M.
London
,
D. P.
Hoytema van Konijnenburg
,
T.
Rendon
,
T.
Buch
,
H. M.
Silva
,
J. J.
Lafaille
,
B. S.
Reis
,
D.
Mucida
.
2016
.
Tissue adaptation of regulatory and intraepithelial CD4+ T cells controls gut inflammation.
Science
352
:
1581
1586
.
5.
He
S.
,
F.
Kahles
,
S.
Rattik
,
M.
Nairz
,
C. S.
McAlpine
,
A.
Anzai
,
D.
Selgrade
,
A. M.
Fenn
,
C. T.
Chan
,
J. E.
Mindur
, et al
2019
.
Gut intraepithelial T cells calibrate metabolism and accelerate cardiovascular disease.
Nature
566
:
115
119
.
6.
Guy-Grand
D.
,
P.
Vassalli
.
2002
.
Gut intraepithelial lymphocyte development.
Curr. Opin. Immunol.
14
:
255
259
.
7.
Guy-Grand
D.
,
O.
Azogui
,
S.
Celli
,
S.
Darche
,
M. C.
Nussenzweig
,
P.
Kourilsky
,
P.
Vassalli
.
2003
.
Extrathymic T cell lymphopoiesis: ontogeny and contribution to gut intraepithelial lymphocytes in athymic and euthymic mice.
J. Exp. Med.
197
:
333
341
.
8.
Lambolez
F.
,
M.
Kronenberg
,
H.
Cheroutre
.
2007
.
Thymic differentiation of TCR alpha beta(+) CD8 alpha alpha(+) IELs.
Immunol. Rev.
215
:
178
188
.
9.
Leishman
A. J.
,
L.
Gapin
,
M.
Capone
,
E.
Palmer
,
H. R.
MacDonald
,
M.
Kronenberg
,
H.
Cheroutre
.
2002
.
Precursors of functional MHC class I- or class II-restricted CD8alphaalpha(+) T cells are positively selected in the thymus by agonist self-peptides.
Immunity
16
:
355
364
.
10.
Baldwin
T. A.
,
K. A.
Hogquist
,
S. C.
Jameson
.
2004
.
The fourth way? Harnessing aggressive tendencies in the thymus.
J. Immunol.
173
:
6515
6520
.
11.
Stritesky
G. L.
,
S. C.
Jameson
,
K. A.
Hogquist
.
2012
.
Selection of self-reactive T cells in the thymus.
Annu. Rev. Immunol.
30
:
95
114
.
12.
Ruscher
R.
,
K. A.
Hogquist
.
2019
.
Development, ontogeny, and maintenance of TCRαβ+ CD8αα IEL.
Curr. Opin. Immunol.
58
:
83
88
.
13.
Verstichel
G.
,
D.
Vermijlen
,
L.
Martens
,
G.
Goetgeluk
,
M.
Brouwer
,
N.
Thiault
,
Y.
Van Caeneghem
,
S.
De Munter
,
K.
Weening
,
S.
Bonte
, et al
2017
.
The checkpoint for agonist selection precedes conventional selection in human thymus.
Sci. Immunol.
2
:
eaah4232
.
14.
Grandjean
C. L.
,
N.
Sumaria
,
S.
Martin
,
D. J.
Pennington
.
2017
.
Increased TCR signal strength in DN thymocytes promotes development of gut TCRαβ(+)CD8αα(+) intraepithelial lymphocytes.
Sci. Rep.
7
:
10659
.
15.
Eberl
G.
,
D. R.
Littman
.
2004
.
Thymic origin of intestinal alphabeta T cells revealed by fate mapping of RORgammat+ cells.
Science
305
:
248
251
.
16.
Gangadharan
D.
,
F.
Lambolez
,
A.
Attinger
,
Y.
Wang-Zhu
,
B. A.
Sullivan
,
H.
Cheroutre
.
2006
.
Identification of pre- and postselection TCRalphabeta+ intraepithelial lymphocyte precursors in the thymus.
Immunity
25
:
631
641
.
17.
McGargill
M. A.
,
K. A.
Hogquist
.
1999
.
Antigen-induced coreceptor down-regulation on thymocytes is not a result of apoptosis.
J. Immunol.
162
:
1237
1245
.
18.
McDonald
B. D.
,
J. J.
Bunker
,
S. A.
Erickson
,
M.
Oh-Hora
,
A.
Bendelac
.
2015
.
Crossreactive αβ T cell receptors are the predominant targets of thymocyte negative selection.
Immunity
43
:
859
869
.
19.
Yamagata
T.
,
D.
Mathis
,
C.
Benoist
.
2004
.
Self-reactivity in thymic double-positive cells commits cells to a CD8 alpha alpha lineage with characteristics of innate immune cells.
Nat. Immunol.
5
:
597
605
.
20.
McDonald
B. D.
,
J. J.
Bunker
,
I. E.
Ishizuka
,
B.
Jabri
,
A.
Bendelac
.
2014
.
Elevated T cell receptor signaling identifies a thymic precursor to the TCRαβ(+)CD4(-)CD8β(-) intraepithelial lymphocyte lineage.
Immunity
41
:
219
229
.
21.
Nakase
K.
,
K.
Kita
,
H.
Anazawa
,
K.
Hoshino
,
S.
Shirakawa
,
I.
Tanaka
,
M.
Tsudo
.
1994
.
Induction of interleukin-2 receptor alpha chain expression of immature acute myelocytic leukemia cells.
Leuk. Res.
18
:
269
274
.
22.
Ruscher
R.
,
R. L.
Kummer
,
Y. J.
Lee
,
S. C.
Jameson
,
K. A.
Hogquist
.
2017
.
CD8αα intraepithelial lymphocytes arise from two main thymic precursors.
Nat. Immunol.
18
:
771
779
.
23.
Golec
D. P.
,
R. E.
Hoeppli
,
L. M.
Henao Caviedes
,
J.
McCann
,
M. K.
Levings
,
T. A.
Baldwin
.
2017
.
Thymic progenitors of TCRαβ+ CD8αα intestinal intraepithelial lymphocytes require RasGRP1 for development.
J. Exp. Med.
214
:
2421
2435
.
24.
Klose
C. S.
,
K.
Blatz
,
Y.
d’Hargues
,
P. P.
Hernandez
,
M.
Kofoed-Nielsen
,
J. F.
Ripka
,
K.
Ebert
,
S. J.
Arnold
,
A.
Diefenbach
,
E.
Palmer
,
Y.
Tanriver
.
2014
.
The transcription factor T-bet is induced by IL-15 and thymic agonist selection and controls CD8αα(+) intraepithelial lymphocyte development.
Immunity
41
:
230
243
.
25.
Klose
C. S. N.
,
J. F.
Hummel
,
L.
Faller
,
Y.
d’Hargues
,
K.
Ebert
,
Y.
Tanriver
.
2018
.
A committed postselection precursor to natural TCRαβ+ intraepithelial lymphocytes.
Mucosal Immunol.
11
:
333
344
.
26.
Ruscher
R.
,
S. T.
Lee
,
O. C.
Salgado
,
E. R.
Breed
,
S. H.
Osum
,
K. A.
Hogquist
.
2020
.
Intestinal CD8αα IELs derived from two distinct thymic precursors have staggered ontogeny.
J. Exp. Med.
217
:
e20192336
.
27.
O’Reilly
L. A.
,
E. A.
Kruse
,
H.
Puthalakath
,
P. N.
Kelly
,
T.
Kaufmann
,
D. C.
Huang
,
A.
Strasser
.
2009
.
MEK/ERK-mediated phosphorylation of Bim is required to ensure survival of T and B lymphocytes during mitogenic stimulation.
J. Immunol.
183
:
261
269
.
28.
Cheng
E. H.
,
M. C.
Wei
,
S.
Weiler
,
R. A.
Flavell
,
T. W.
Mak
,
T.
Lindsten
,
S. J.
Korsmeyer
.
2001
.
BCL-2, BCL-X(L) sequester BH3 domain-only molecules preventing BAX- and BAK-mediated mitochondrial apoptosis.
Mol. Cell
8
:
705
711
.
29.
Bouillet
P.
,
J. F.
Purton
,
D. I.
Godfrey
,
L. C.
Zhang
,
L.
Coultas
,
H.
Puthalakath
,
M.
Pellegrini
,
S.
Cory
,
J. M.
Adams
,
A.
Strasser
.
2002
.
BH3-only Bcl-2 family member Bim is required for apoptosis of autoreactive thymocytes. [Published erratum appears in 2002 Nature 618: 108.]
Nature
415
:
922
926
.
30.
Li
K. P.
,
A.
Fähnrich
,
E.
Roy
,
C. M.
Cuda
,
H. L.
Grimes
,
H. R.
Perlman
,
K.
Kalies
,
D. A.
Hildeman
.
2017
.
Temporal expression of Bim limits the development of agonist-selected thymocytes and skews their TCRβ repertoire.
J. Immunol.
198
:
257
269
.
31.
Wojciechowski
S.
,
M. B.
Jordan
,
Y.
Zhu
,
J.
White
,
A. J.
Zajac
,
D. A.
Hildeman
.
2006
.
Bim mediates apoptosis of CD127(lo) effector T cells and limits T cell memory.
Eur. J. Immunol.
36
:
1694
1706
.
32.
Wojciechowski
S.
,
P.
Tripathi
,
T.
Bourdeau
,
L.
Acero
,
H. L.
Grimes
,
J. D.
Katz
,
F. D.
Finkelman
,
D. A.
Hildeman
.
2007
.
Bim/Bcl-2 balance is critical for maintaining naive and memory T cell homeostasis.
J. Exp. Med.
204
:
1665
1675
.
33.
Kurtulus
S.
,
P.
Tripathi
,
M. E.
Moreno-Fernandez
,
A.
Sholl
,
J. D.
Katz
,
H. L.
Grimes
,
D. A.
Hildeman
.
2011
.
Bcl-2 allows effector and memory CD8+ T cells to tolerate higher expression of Bim.
J. Immunol.
186
:
5729
5737
.
34.
Wirasinha
R. C.
,
A.
Chan
,
J. Y.
Yap
,
D. Y.
Hu
,
C. E.
Teh
,
D. H. D.
Gray
,
C. C.
Goodnow
,
S. R.
Daley
.
2019
.
Deletion of self-reactive CCR7- thymocytes in the absence of MHC expression on thymic epithelial cells.
Cell Death Differ.
26
:
2727
2739
.
35.
Pobezinsky
L. A.
,
G. S.
Angelov
,
X.
Tai
,
S.
Jeurling
,
F.
Van Laethem
,
L.
Feigenbaum
,
J. H.
Park
,
A.
Singer
.
2012
.
Clonal deletion and the fate of autoreactive thymocytes that survive negative selection.
Nat. Immunol.
13
:
569
578
.
36.
Lai
Y. G.
,
M. S.
Hou
,
A.
Lo
,
S. T.
Huang
,
Y. W.
Huang
,
H. F.
Yang-Yen
,
N. S.
Liao
.
2013
.
IL-15 modulates the balance between Bcl-2 and Bim via a Jak3/1-PI3K-Akt-ERK pathway to promote CD8αα+ intestinal intraepithelial lymphocyte survival.
Eur. J. Immunol.
43
:
2305
2316
.
37.
Nakayama
K.
,
K.
Nakayama
,
I.
Negishi
,
K.
Kuida
,
H.
Sawa
,
D. Y.
Loh
.
1994
.
Targeted disruption of Bcl-2 alpha beta in mice: occurrence of gray hair, polycystic kidney disease, and lymphocytopenia.
Proc. Natl. Acad. Sci. USA
91
:
3700
3704
.
38.
Raynor
J.
,
A.
Sholl
,
D. R.
Plas
,
P.
Bouillet
,
C. A.
Chougnet
,
D. A.
Hildeman
.
2013
.
IL-15 fosters age-driven regulatory T cell accrual in the face of declining IL-2 levels.
Front. Immunol.
4
:
161
.
39.
Herold
M. J.
,
R.
Stuchbery
,
D.
Mérino
,
T.
Willson
,
A.
Strasser
,
D.
Hildeman
,
P.
Bouillet
.
2014
.
Impact of conditional deletion of the pro-apoptotic BCL-2 family member BIM in mice.
Cell Death Dis.
5
:
e1446
.
40.
Lefrancois
L.
,
N.
Lycke
.
2001
.
Isolation of mouse small intestinal intraepithelial lymphocytes, Peyer’s patch, and lamina propria cells.
Curr. Protoc. Immunol.
Chapter 3: Unit 3.19
.
41.
Thorp
E.
,
Y.
Li
,
L.
Bao
,
P. M.
Yao
,
G.
Kuriakose
,
J.
Rong
,
E. A.
Fisher
,
I.
Tabas
.
2009
.
Brief report: increased apoptosis in advanced atherosclerotic lesions of Apoe-/- mice lacking macrophage Bcl-2.
Arterioscler. Thromb. Vasc. Biol.
29
:
169
172
.
42.
Kurd
N. S.
,
A.
Hoover
,
J.
Yoon
,
B. M.
Weist
,
L.
Lutes
,
S. W.
Chan
,
E. A.
Robey
.
2021
.
Factors that influence the thymic selection of CD8αα intraepithelial lymphocytes.
Mucosal Immunol.
14
:
68
79
.
43.
Guo
X.
,
Y.
Tanaka
,
M.
Kondo
.
2015
.
Thymic precursors of TCRαβ(+)CD8αα(+) intraepithelial lymphocytes are negative for CD103.
Immunol. Lett.
163
:
40
48
.
44.
Sato
T.
,
S.
Ohno
,
T.
Hayashi
,
C.
Sato
,
K.
Kohu
,
M.
Satake
,
S.
Habu
.
2005
.
Dual functions of Runx proteins for reactivating CD8 and silencing CD4 at the commitment process into CD8 thymocytes.
Immunity
22
:
317
328
.
45.
Moran
A. E.
,
K. L.
Holzapfel
,
Y.
Xing
,
N. R.
Cunningham
,
J. S.
Maltzman
,
J.
Punt
,
K. A.
Hogquist
.
2011
.
T cell receptor signal strength in Treg and iNKT cell development demonstrated by a novel fluorescent reporter mouse.
J. Exp. Med.
208
:
1279
1289
.
46.
Akbar
A. N.
,
N. J.
Borthwick
,
R. G.
Wickremasinghe
,
P.
Panayoitidis
,
D.
Pilling
,
M.
Bofill
,
S.
Krajewski
,
J. C.
Reed
,
M.
Salmon
.
1996
.
Interleukin-2 receptor common gamma-chain signaling cytokines regulate activated T cell apoptosis in response to growth factor withdrawal: selective induction of anti-apoptotic (bcl-2, bcl-xL) but not pro-apoptotic (bax, bcl-xS) gene expression.
Eur. J. Immunol.
26
:
294
299
.
47.
Zhu
Y.
,
B. J.
Swanson
,
M.
Wang
,
D. A.
Hildeman
,
B. C.
Schaefer
,
X.
Liu
,
H.
Suzuki
,
K.
Mihara
,
J.
Kappler
,
P.
Marrack
.
2004
.
Constitutive association of the proapoptotic protein Bim with Bcl-2-related proteins on mitochondria in T cells.
Proc. Natl. Acad. Sci. USA
101
:
7681
7686
.
48.
Chopp
L. B.
,
V.
Gopalan
,
T.
Ciucci
,
A.
Ruchinskas
,
Z.
Rae
,
M.
Lagarde
,
Y.
Gao
,
C.
Li
,
M.
Bosticardo
,
F.
Pala
, et al
2020
.
An integrated epigenomic and transcriptomic map of mouse and human αβ T cell development.
Immunity
53
:
1182
1201.e8
.
49.
Oh-Hora
M.
,
N.
Komatsu
,
M.
Pishyareh
,
S.
Feske
,
S.
Hori
,
M.
Taniguchi
,
A.
Rao
,
H.
Takayanagi
.
2013
.
Agonist-selected T cell development requires strong T cell receptor signaling and store-operated calcium entry.
Immunity
38
:
881
895
.
50.
Lauritsen
J. P.
,
S.
Kurella
,
S. Y.
Lee
,
J. M.
Lefebvre
,
M.
Rhodes
,
J.
Alberola-Ila
,
D. L.
Wiest
.
2008
.
Egr2 is required for Bcl-2 induction during positive selection.
J. Immunol.
181
:
7778
7785
.
51.
Lai
Y. G.
,
M. S.
Hou
,
Y. W.
Hsu
,
C. L.
Chang
,
Y. H.
Liou
,
M. H.
Tsai
,
F.
Lee
,
N. S.
Liao
.
2008
.
IL-15 does not affect IEL development in the thymus but regulates homeostasis of putative precursors and mature CD8 alpha alpha+ IELs in the intestine.
J. Immunol.
180
:
3757
3765
.
52.
Lambolez
F.
,
O.
Azogui
,
A. M.
Joret
,
C.
Garcia
,
H.
von Boehmer
,
J.
Di Santo
,
S.
Ezine
,
B.
Rocha
.
2002
.
Characterization of T cell differentiation in the murine gut.
J. Exp. Med.
195
:
437
449
.
53.
Konkel
J. E.
,
T.
Maruyama
,
A. C.
Carpenter
,
Y.
Xiong
,
B. F.
Zamarron
,
B. E.
Hall
,
A. B.
Kulkarni
,
P.
Zhang
,
R.
Bosselut
,
W.
Chen
.
2011
.
Control of the development of CD8αα+ intestinal intraepithelial lymphocytes by TGF-β.
Nat. Immunol.
12
:
312
319
.
54.
McKean
D. J.
,
C. J.
Huntoon
,
M. P.
Bell
,
X.
Tai
,
S.
Sharrow
,
K. E.
Hedin
,
A.
Conley
,
A.
Singer
.
2001
.
Maturation versus death of developing double-positive thymocytes reflects competing effects on Bcl-2 expression and can be regulated by the intensity of CD28 costimulation.
J. Immunol.
166
:
3468
3475
.
55.
Gray
D. H.
,
F.
Kupresanin
,
S. P.
Berzins
,
M. J.
Herold
,
L. A.
O’Reilly
,
P.
Bouillet
,
A.
Strasser
.
2012
.
The BH3-only proteins Bim and Puma cooperate to impose deletional tolerance of organ-specific antigens.
Immunity
37
:
451
462
.
56.
Erlacher
M.
,
V.
Labi
,
C.
Manzl
,
G.
Böck
,
A.
Tzankov
,
G.
Häcker
,
E.
Michalak
,
A.
Strasser
,
A.
Villunger
.
2006
.
Puma cooperates with Bim, the rate-limiting BH3-only protein in cell death during lymphocyte development, in apoptosis induction.
J. Exp. Med.
203
:
2939
2951
.
57.
Williams
O.
,
T.
Norton
,
M.
Halligey
,
D.
Kioussis
,
H. J.
Brady
.
1998
.
The action of Bax and bcl-2 on T cell selection.
J. Exp. Med.
188
:
1125
1133
.
58.
Reis
B. S.
,
D. P.
Hoytema van Konijnenburg
,
S. I.
Grivennikov
,
D.
Mucida
.
2014
.
Transcription factor T-bet regulates intraepithelial lymphocyte functional maturation.
Immunity
41
:
244
256
.
59.
Wirasinha
R. C.
,
M.
Singh
,
S. K.
Archer
,
A.
Chan
,
P. F.
Harrison
,
C. C.
Goodnow
,
S. R.
Daley
.
2018
.
αβ T-cell receptors with a central CDR3 cysteine are enriched in CD8αα intraepithelial lymphocytes and their thymic precursors.
Immunol. Cell Biol.
96
:
553
561
.
60.
Lutz
C.
,
M.
Mozaffari
,
V.
Tosevski
,
M.
Caj
,
P.
Cippà
,
B. L.
McRae
,
C. L.
Graff
,
G.
Rogler
,
M.
Fried
,
M.
Hausmann
.
2015
.
Increased lymphocyte apoptosis in mouse models of colitis upon ABT-737 treatment is dependent upon BIM expression.
Clin. Exp. Immunol.
181
:
343
356
.

The authors have no financial conflicts of interest.

Supplementary data