The thymus generates two divergent types of lymphocytes, innate and adaptive T cells. Innate T cells such as invariant NKT cells provide immediate immune defense, whereas adaptive T cells require a phase of expansion and functional differentiation outside the thymus. Naive adaptive T lymphocytes should not proliferate much after positive selection in the thymus to ensure a highly diverse TCR repertoire. In contrast, oligoclonal innate lymphocyte populations are efficiently expanded through intrathymic proliferation. For CD4+Foxp3+ regulatory T cells (Tregs), which are thought to be generated by agonist recognition, it is not clear whether they proliferate upon thymic selection. In this study, we investigated thymic and peripheral T cell proliferation by genetic pulse labeling. To this end, we used a mouse model in which all developing αβ thymocytes were marked by expression of a histone 2B–enhanced GFP (H2BeGFP) fusion-protein located within the Tcrd locus (TcrdH2BeGFP). This reporter gene was excised during TCR α-chain VJ-recombination, and the retained H2BeGFP signal was thus diluted upon cell proliferation. We found that innate T cells such as CD1d-restricted invariant NKT cells all underwent a phase of intense intrathymic proliferation, whereas adaptive CD4+ and CD8+ single-positive thymocytes including thymic Tregs cycled, on average, only once after final selection. After thymic exit, retention or loss of very stable H2BeGFP signal indicated the proliferative history of peripheral αβ T cells. There, peripheral Tregs showed lower levels of H2BeGFP compared with CD4+Foxp3 T cells. This further supports the hypothesis that the Treg repertoire is shaped by self-Ag recognition in the steady-state.

This article is featured in In This Issue, p.2023

Proliferation of αβ T cells occurs at different stages during their development and differentiation. The earliest CD4CD8 double-negative (DN) T cell precursors are proliferative, presumably to compensate for the few thymus seeding progenitors that can enter the thymus at a time (1, 2). Next, at the CD4CD8 DN CD25+ stage, selection of productively rearranged Tcrb genes induces a phase of strong proliferation that is called β-selection and serves to expand clones expressing a TCR β-chain (3), and that is also required for αβ-lineage commitment and progression to the CD4+CD8+ double-positive (DP) stage (4). Then, proliferation is terminated before Tcra rearrangement and positive selection. It is not clear to what extent DP thymocytes divide after positive selection of their αβ TCR. In theory, the lack of further proliferation after final αβ thymocyte selection would maximize the diversity of αβ T cells. However, it is conceivable that a few intrathymic cell divisions support efficient generation of functional T cells whereas maintaining a still highly diverse TCR repertoire (5, 6). For late DP and single-positive (SP) thymocytes, eventual encounter of agonistic TCR ligands can either lead to negative selection by apoptosis, thereby ensuring central tolerance, or induce survival by a process called agonist selection. The best studied agonist selection is the differentiation of invariant NKT (iNKT) cells (7). After recognition of instructive signals during TCR engagement by CD1d-expressing cortical thymocytes, intrathymic proliferation generates large numbers of oligoclonal innate T cell populations with limited TCR diversity (8). In contrast, the role of agonist selection for thymic differentiation of Foxp3+ regulatory T cells (Tregs) is currently controversial (912), and it is not known whether self-Ag recognition induces intrathymic proliferation of Tregs.

There are numerous methods for measurement of cell proliferation, including quantification of cells that incorporated [3H]thymidine or BrdU (13, 14), CFSE labeling (15), or direct detection of recently divided cells by Ki-67 expression (16). Earlier studies using these methods suggested intrathymic expansion of mature thymocytes before export to the periphery (17). More recently, BAC-transgenic mice with multiple copies of GFP under the control of the Rag2 promoter (18) were used as a genetic tool to monitor the dynamics of thymic maturation, emigration (19), and to directly identify recent thymic emigrants (RTEs) (20, 21). In this Rag2p-GFP system, GFP-reporter transcription terminates after Tcra rearrangement in DP cells, but GFP fluorescence is detectable in peripheral T cells because of the considerable half-life of GFP in mouse T cells estimated to range between 16 and 18 h (20) and 54 and 56 h (19). Thus, the observation that the GFP signal in Rag2p-GFP mice lingers in SP thymocytes and RTEs further supports the idea that conventional αβ T cells do not cycle much after positive selection.

In this study, we used a genetic system in which a histone 2B–enhanced GFP (H2BeGFP) was introduced into the 3′ untranslated region of the Tcrd constant gene (22). In these TcrdH2BeGFP mice, all γδ T cells are marked by high expression of a very stable H2BeGFP fusion-protein. In addition, the reporter gene is transcribed in all DN thymocytes but excised during Vα-Jα recombination in DP thymocytes. Because the estimated half-life of H2BeGFP is in the range of several months (23), it is retained in nonproliferating postselection αβ T cells and allows to track their proliferative history in vivo. Using this genetic system for pulsing DP thymocytes, we monitored the differential dynamics of conventional T cells, Foxp3+ Tregs, and iNKT cells. We observed that all developing iNKT cells completely lost H2BeGFP fluorescence at an early developmental stage (8), thereby validating the proliferation-reporter system. In contrast, CD4+ and CD8+ SP cells still displayed intermediate H2BeGFP levels after thymic maturation and, therefore, had proliferated only to a minor degree. After thymic exit, the sole RTE showing proliferation were CD8+, which was in line with results from Fink and colleagues (20). In the periphery, resting, naive CD4+ or CD8+ T cells continued to retain an intermediate H2BeGFP signal. Even massive Ag-specific T cell immune responses to viral infection did not induce bystander proliferation among the peripheral naive T cell pool. Rather, homeostatic proliferation of naive T cells was observed only under lymphopenic conditions, for example, after adoptive transfer into Rag−/−γc−/− recipients. Furthermore, nuclear localization of the H2BeGFP reporter within the chromatin facilitated simultaneous detection of intracellular Foxp3 protein and monitoring of proliferative history. Although Foxp3+CD4+ thymocytes and mature Foxp3CD4+ and CD8+ SP cells showed similar intrathymic proliferation, Foxp3+ Tregs in the periphery displayed lower H2BeGFP levels than Foxp3CD4+ T cells. This was consistent with the hypothesis that Tregs proliferate after recognizing self-Ag in the periphery and undergo organ-specific shaping of their TCR repertoire (24, 25). In summary, we present new data on the different postselection proliferation dynamics of innate αβ T cells such as iNKT cells as compared with adaptive αβ T cell populations, which include Foxp3+ Tregs.

TcrdH2BeGFP mice have been described previously (22). C57BL/6J mice were obtained from Charles River (Germany). Mice were bred and housed under specific pathogen-free conditions in the central animal facilities of Hannover Medical School or Centre d’Immunology de Marseille-Luminy and used at 6–12 wk of age. All experiments were approved by local institutional animal care and research committees, and authorized by local governments. This study was conducted in accordance with the European Communities Council Directive86/609/EEC for the protection of animals used for experimental purposes.

Before staining, cells were incubated at 4°C for 5 min with Fc-blocking reagent (mAb 2.4G2). For multicolor staining, cells were incubated with conjugate Abs for 15 min at 4°C, washed with PBS/3% FCS, and, if required, incubated with secondary Ab streptavidin-PE-Cy7 for 10 min at 4°C and washed with PBS/3% FCS. For CCR7 mAb staining, cells were incubated with anti–CCR7-PE Ab in T cell medium for 30 min at 37°C, washed with PBS/3% FCS, and incubated with surface Abs for 15 min at 4°C. Intracellular Foxp3 staining was performed using Foxp3 staining buffer (eBioscience) according to the manufacturer’s protocol. The same buffer system was used for staining of intracellular helios with anti-helios mAb clone 22F6 conjugated with Alexa Fluor 647 (Biolegend). For Neuropilin-1 (Nrp1) staining, we used biotinylated polyclonal goat anti-Nrp1 IgG (R&D Systems), followed by PE-Cy7–conjugated streptavidin. In some experiments, thymic iNKT cells were enriched by labeling the cells with anti–CD1d/α-GalCer tetramer-allophycocyanin and subsequent magnetic enrichment with anti–allophycocyanin-Microbeads (Miltenyi). Flow cytometric analysis was performed on a LSR cytometer (BD Bioscience) equipped with the BD FACSDiva software. Because ∼1% of all αβ thymocytes show monoallelic Tcra rearrangement, a fraction of those retained H2BeGFP levels comparable with γδ T cells (unpublished data). Therefore, we excluded H2BeGFPhigh CD4+ and CD8+ T cells from our analysis. The validity of this approach was confirmed by analysis of F1 offspring from Tcrd-H2BeGFP reporter mice crossed to Tcra-deficient mice as described previously (26). In experiments where data are shown as fold change of H2BeGFP fluorescence intensity, we always included separate analyses of the respective cell populations from nonfluorescent C57BL/6 control mice to calculate relative values compared with WT control. Further gating strategies are described in the figure legends. Postacquisition analyses were conducted using the FlowJo software (TreeStar).

For intrathymic injection of d-biotin-N-hydroxysuccinimide ester (biotin-OSu; Anaspec) to mark RTEs, mice were anesthetized and 10 μl biotin-OSu solution (2.5 μg/μl in DMSO/PBS) was injected into one thymic lobe. All mice were analyzed after 20 h. Biotin+ cells were identified by streptavidin-Cy5 conjugation.

To investigate homeostatic proliferation under lymphopenic conditions, we adoptively transferred 5 × 107TcrdH2BeGFP lymphocytes derived from filtered peripheral lymph node suspensions into congenic Rag−/−γc−/− (C57BL/6) recipients via the lateral tail vein. For thymectomy, 4- to 6-wk-old TcrdH2BeGFP mice were anesthetized and thymi removed by suction as described elsewhere (27).

Infections were initiated by i.p. delivery of 5 × 104 PFU thawed murine CMV (MCMV) stocks (28). Smith strain salivary gland extracts were prepared as described previously (29). Infections were initiated on different days so that each animal in a given experiment was sacrificed on the same day.

Group data were compared with one-way ANOVA or repeated-measures ANOVA with Tukey posttest (GraphPad Prism) to generate p values.

Developing thymocytes open the Tcrd locus at the transition from DN1b to DN2 stage, regardless of their final commitment to αβ or γδ lineage (22). Later at the CD4+CD8+ DP stage, the Tcrd locus is excised during Vα to Jα rearrangement (30). Thus, in TcrdH2BeGFP mice, the H2BeGFP reporter gene within the Tcrd locus is transiently transcribed during CD4CD8 DN to CD4+CD8+ DP stages, but removed during Tcra rearrangement (Supplemental Fig. 1A). After positive selection and excision of the reporter gene, the very stable H2BeGFP is retained and legacy fluorescence is thus distributed among daughter cells upon proliferation (Supplemental Fig. 1B). To control that loss of H2BeGFP correlated with cell proliferation, we stimulated T cells derived from spleen and peripheral lymph nodes in an in vitro proliferation assay and double-checked with BrdU staining. As shown in Supplemental Fig. 2A, in vivo proliferated BrdU+ CD4+ and CD8+ T cells possessed significantly lower H2BeGFP levels compared with BrdU cells. These data demonstrate a direct link between proliferation and H2BeGFP loss. To exclude that TCR rearrangement excision circles might still drive H2BeGFP expression in αβ T cells, we analyzed Tcrd transcription of sorted H2BeGFP and H2BeGFP+ αβ T cells. No transcription of rearranged Tcrd genes was detected, suggesting that H2BeGFP+ αβ T cells do not transcribe H2BeGFP mRNA (Supplemental Fig. 2B). Next, we monitored how dilution of eFluor 670 staining correlated to loss of H2BeGFP signal to determine how many cell divisions could be detected by dilution of H2BeGFP in vitro. After anti-CD3/anti-CD28 stimulation, peripheral CD4+ and CD8+ T cells showed complete loss of H2BeGFP fluorescence after approximately two cell cycles (Supplemental Fig. 2C). Therefore, retention of H2BeGFP signal identifies peripheral αβ T cells that have undergone less than two cell divisions after exit from the thymus.

Within the thymus, we observed even more intense labeling of all DP developing αβ T cells as compared with peripheral T cells. DP green fluorescence could be quantified as 30-fold of the background in control C57BL/6 mice (fold median H2BeGFP fluorescence intensity [MFI] of C57BL/6 control throughout the article) in DP thymocytes (Fig. 1). This qualified DP thymocytes as H2BeGFPintermediate with regard to the ∼10-fold higher fluorescence observed in γδ T cells. Thymic CD4+ and CD8+ SP T cells showed H2BeGFP signal of 15- and 10-fold MFI, respectively (Fig. 1). Because signal intensity principally decreases through cell proliferation, 50% loss of fluorescence between DP and SP stage is consistent with the occurrence, on average, of one cell division.

FIGURE 1.

Thymic αβ T cells retain intermediate H2BeGFP fluorescence in TcrdH2BeGFP mice. Histograms show the H2BeGFP level, and graph shows quantification H2BeGFP MFI fold change of γδ (CD4CD8GL3+H2BeGFPhigh), DP (CD4+CD8+), CD4SP (CD4+CD8CD25NK1.1), and CD8SP (CD8+CD4CD25NK1.1) T cells in unchallenged TcrdH2BeGFP mice compared with C57BL/6 controls (gray filled histograms). Representative and combined data from three different experiments (n = 15) are shown.

FIGURE 1.

Thymic αβ T cells retain intermediate H2BeGFP fluorescence in TcrdH2BeGFP mice. Histograms show the H2BeGFP level, and graph shows quantification H2BeGFP MFI fold change of γδ (CD4CD8GL3+H2BeGFPhigh), DP (CD4+CD8+), CD4SP (CD4+CD8CD25NK1.1), and CD8SP (CD8+CD4CD25NK1.1) T cells in unchallenged TcrdH2BeGFP mice compared with C57BL/6 controls (gray filled histograms). Representative and combined data from three different experiments (n = 15) are shown.

Close modal

DP thymocytes require (at least weak) stimulation of their TCR to pass positive selection. However, it is unclear whether such TCR signals induce cell proliferation. We observed lower H2BeGFP levels in TCRhighCD69+ (postselection) than in TCRlowCD69 (preselection) DP thymocytes (Fig. 2). This suggests that thymocytes proliferate during positive selection. To further distinguish different maturation steps of CD4+ and CD8+ SP thymocytes, we gated them into CD24+CD69+ (semimature) and CD24CD69 (mature) thymocytes. We observed that CD4+ and CD8+ SP cells lost further H2BeGFP signal during thymic maturation (Fig. 2). Together, these observations provide fresh evidence that thymocytes undergo one to two cell divisions after positive selection. Such duplication of the naive T cell pool may warrant a trade-off between maximal TCR repertoire diversity and the presence of more than one representative of a crucial T cell clone. Our results are in line with previous studies that showed late intrathymic proliferation of SP T cells before thymic exit (5, 6, 17).

FIGURE 2.

Proliferation of postselection thymocytes quantified by loss of H2BeGFP. CD4+CD8+ DP thymocytes before and after positive selection were gated as CD69TCRβlow and CD69+TCRβhigh, respectively (upper left panel). During selection, CD4+ and CD8+ SP T cells lose CD24 and CD69 expression; therefore, CD24+CD69+ SP T cells were defined as semimature and CD24CD69 SP T cells as mature. Graph shows fold change of H2BeGFP MFI of the indicated thymocyte populations compared with C57BL/6 mice. Combined data from three different experiments (n = 12) are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest. ***p < 0.001.

FIGURE 2.

Proliferation of postselection thymocytes quantified by loss of H2BeGFP. CD4+CD8+ DP thymocytes before and after positive selection were gated as CD69TCRβlow and CD69+TCRβhigh, respectively (upper left panel). During selection, CD4+ and CD8+ SP T cells lose CD24 and CD69 expression; therefore, CD24+CD69+ SP T cells were defined as semimature and CD24CD69 SP T cells as mature. Graph shows fold change of H2BeGFP MFI of the indicated thymocyte populations compared with C57BL/6 mice. Combined data from three different experiments (n = 12) are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest. ***p < 0.001.

Close modal

Next, we asked whether T cells continued to proliferate in the periphery after thymic exit. To this end, TcrdH2BeGFP mice were intrathymically injected with a reactive biotin derivative, and peripheral T cells were analyzed 20 h later. At this time point, 0.30 ± 0.19% of peripheral CD4+ T cells and 0.23 ± 0.15% of peripheral CD8+ T cells were identified as RTEs by surface biotin labeling (Fig. 3). Compared with non-RTEs, RTEs displayed slightly higher H2BeGFP levels. Furthermore, CD8+ RTEs, but not CD4+ RTEs, showed lower H2BeGFP intensity in comparison with mature SP thymocytes, indicative of fewer turnovers after or during thymic exit (Fig. 3). To address how H2BeGFP levels were distributed among peripheral T cell subsets in steady-state, we correlated TcrdH2BeGFP fluorescence with expression of T cell activation markers. l-selectin, also known as CD62L, is a cell adhesion molecule that facilitates entry of naive and memory T cells into the lymph node via high endothelial venules. In contrast, high expression of CD44, which is a receptor for intracellular matrix proteins, is a hallmark of activated effector T cells. Accordingly, when gating CD4+ and CD8+ T cells in relation to their expression of CD62L and CD44, we observed higher H2BeGFP levels in naive CD62L+CD44intCD4+ and CD62L+CD44lowCD8+ T cells as compared with activated CD44+ effector T cells, which completely lost H2BeGFP intensity (Fig. 4A). Also, CD62L+CD44high CD8+ memory T cells had a fully proliferated phenotype (Fig. 4A). Furthermore, H2BeGFP levels correlated negatively with expression of the immediate early activation marker CD69 (Fig. 4B). In contrast, the lymph node homing receptor CCR7 is expressed on naive T cells and on memory CD8+ T cells, and accordingly showed a positive correlation to H2BeGFP levels in CD4+, but not CD8+ T cells (Fig. 4C).

FIGURE 3.

CD8+ RTEs, but not CD4+ RTEs, proliferate upon thymic exit. Twenty hours after intrathymic injection of reactive biotin, CD62L+biotin+ T cells in peripheral lymph nodes were identified and gated as RTEs. Right panels show fold change of H2BeGFP MFI of the indicated CD4+ (upper panels) or CD8+ (lower panels) T cells compared with C57BL/6 control mice. Combined data from two different experiments (n = 6–8) are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest. ***p < 0.001.

FIGURE 3.

CD8+ RTEs, but not CD4+ RTEs, proliferate upon thymic exit. Twenty hours after intrathymic injection of reactive biotin, CD62L+biotin+ T cells in peripheral lymph nodes were identified and gated as RTEs. Right panels show fold change of H2BeGFP MFI of the indicated CD4+ (upper panels) or CD8+ (lower panels) T cells compared with C57BL/6 control mice. Combined data from two different experiments (n = 6–8) are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest. ***p < 0.001.

Close modal
FIGURE 4.

Activation of peripheral T cells correlates with loss of H2BeGFP. (A) Histograms show CD62L+CD44 (blue), CD62LCD44 (green), CD62LCD44+ (red), and CD62L+CD44+ (orange) subsets of CD4+ (upper panel) and CD8+ (lower panel) T cells of peripheral lymph nodes from TcrdH2BeGFP mice compared with C57BL/6 (filled gray). Graphs show fold change of H2BeGFP MFI for all different CD4+ (left) and CD8+ (right) T cell subsets compared with C57BL/6 mice. Lower panels show fold change of H2BeGFP MFI of (B) CD69 and CD69+ and (C) CCR7 and CCR7+ subsets of CD4+ and CD8+ T cells derived from TcrdH2BeGFP compared with C57BL/6 mice. Representative and combined data from one [(B) CD69; n = 4] or three [(A, B) CCR7; n = 12] experiments are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest (A) and Student t test (B). ***p < 0.001.

FIGURE 4.

Activation of peripheral T cells correlates with loss of H2BeGFP. (A) Histograms show CD62L+CD44 (blue), CD62LCD44 (green), CD62LCD44+ (red), and CD62L+CD44+ (orange) subsets of CD4+ (upper panel) and CD8+ (lower panel) T cells of peripheral lymph nodes from TcrdH2BeGFP mice compared with C57BL/6 (filled gray). Graphs show fold change of H2BeGFP MFI for all different CD4+ (left) and CD8+ (right) T cell subsets compared with C57BL/6 mice. Lower panels show fold change of H2BeGFP MFI of (B) CD69 and CD69+ and (C) CCR7 and CCR7+ subsets of CD4+ and CD8+ T cells derived from TcrdH2BeGFP compared with C57BL/6 mice. Representative and combined data from one [(B) CD69; n = 4] or three [(A, B) CCR7; n = 12] experiments are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest (A) and Student t test (B). ***p < 0.001.

Close modal

Next, we provided inflammatory or lymphopenic stimuli to monitor peripheral T cell proliferation. First, TcrdH2BeGFP reporter mice were infected with a sublethal dose of MCMV, an infection that is predominantly controlled by CD8+ effector T cells (31, 32). Six days after MCMV infection, we observed a considerably increased pool of activated CD62LCD44+CD8+ and CD62LCD44+CD4+ T cells indicative of a strong primary T cell response (Fig. 5A). We found that all activated CD4+ and CD8+ T cells had lost their H2BeGFP signal; however, this strong response did not induce bystander proliferation in either CD4+ or CD8+ CD62L+CD44int T cells (Fig. 5A). In contrast, we observed loss of H2BeGFP signal, which indicated homeostatic proliferation of CD4+ or CD8+TcrdH2BeGFP T cells after inducing lymphopenia, for example, after thymectomy of TcrdH2BeGFP mice (Fig. 5B) or even faster after adoptive cell transfer into Rag−/−γc−/− recipients (Fig. 5C). Taken together, our data indicate that nonactivated, naive CD4+ and CD8+ T cells show surprisingly little turnover under steady-state and inflammatory conditions, whereas homeostatic proliferation can maintain the peripheral T cell pool if necessary because of lymphopenia.

FIGURE 5.

Lymphopenia but not MCMV infection induces bystander proliferation of naive T cells. (A) Histograms show H2BeGFP levels of naive and activated splenic CD8+ (upper panel) and CD4+ (lower panel) T cells, respectively, at 0, 3, 6, 9, and 12 d after MCMV infection. (B) Four- to 6-wk-old mice were thymectomized, and percentages of H2BeGFP+ CD4+ (left) and CD8+ (right) T cells were monitored weekly in blood samples. (C) A total of 5 × 107 peripheral lymph node cells derived from TcrdH2BeGFP mice was adoptively transferred into Rag−/−γc−/− recipients. The percentage of H2BeGFP+ CD4+ (left) and CD8+ (right) T cells was monitored weekly in blood samples. Representative data from two independent experiments are shown.

FIGURE 5.

Lymphopenia but not MCMV infection induces bystander proliferation of naive T cells. (A) Histograms show H2BeGFP levels of naive and activated splenic CD8+ (upper panel) and CD4+ (lower panel) T cells, respectively, at 0, 3, 6, 9, and 12 d after MCMV infection. (B) Four- to 6-wk-old mice were thymectomized, and percentages of H2BeGFP+ CD4+ (left) and CD8+ (right) T cells were monitored weekly in blood samples. (C) A total of 5 × 107 peripheral lymph node cells derived from TcrdH2BeGFP mice was adoptively transferred into Rag−/−γc−/− recipients. The percentage of H2BeGFP+ CD4+ (left) and CD8+ (right) T cells was monitored weekly in blood samples. Representative data from two independent experiments are shown.

Close modal

In addition to producing a diverse repertoire of adaptive T cells, the thymus generates innate T cell populations such as iNKT with limited TCR diversity. After final TCR rearrangement, iNKT undergo intrathymic proliferation to efficiently produce an oligoclonal T cell population (33). In this study, we investigated intrathymic proliferation during the developmental stages of CD1d-restricted iNKT cells as an example for agonist-selected αβ T cells. Thereby, iNKT cell precursors (stage 0, defined as CD1d/aGC+CD44loCD24hiNK1.1) (8) are thought to be agonist activated by endogenous CD1d-bound lipid ligands, and thus show a strong TCR activation (7, 34). Stage 0 iNKT cells from TcrdH2BeGFP mice were showing fluorescence levels similar to CD4+ SP T cells (Fig. 6). But already at stage 1 (CD1d/aGC+CD44loCD24loNK1.1), all developing iNKT cells had completely lost their H2BeGFP signal, indicating that progression from stage 0 to stage 1 is stringently linked to intrathymic cell division (Fig. 6). Consistently, subsequent iNKT stages 2 and 3 also showed no reappearance of the lost H2BeGFP label. Therefore, iNKT development validated the TcrdH2BeGFP reporter system as a reliable genetic tool for in vivo monitoring of postselection T cell proliferation within the thymus.

FIGURE 6.

iNKT cells vigorously proliferate during thymic development. Thymocytes of TcrdH2BeGFP mice were enriched for CD1d/α-GalCer tetramer (CD1d-Tet)–positive cells using magnetic beads and gated on Tet+TCRβ+ cells. Different developmental iNKT subsets were distinguished by NK1.1, CD44, and CD24 staining (upper panel). Histogram and fold change of MFI of H2BeGFP of stage 0 (red), stage 1 (orange), stage 2 (green), and stage 3 (blue) iNKT cells is shown compared with C57BL/6 control (gray filled). Representative data from two different experiments (n = 11) are shown.

FIGURE 6.

iNKT cells vigorously proliferate during thymic development. Thymocytes of TcrdH2BeGFP mice were enriched for CD1d/α-GalCer tetramer (CD1d-Tet)–positive cells using magnetic beads and gated on Tet+TCRβ+ cells. Different developmental iNKT subsets were distinguished by NK1.1, CD44, and CD24 staining (upper panel). Histogram and fold change of MFI of H2BeGFP of stage 0 (red), stage 1 (orange), stage 2 (green), and stage 3 (blue) iNKT cells is shown compared with C57BL/6 control (gray filled). Representative data from two different experiments (n = 11) are shown.

Close modal

Tregs are another T cell subset that is discussed to be agonist selected (11, 34, 35). Therefore, we asked whether recognition of self-peptides induced proliferation within developing Tregs when compared with naive CD4+Foxp3 T cells or iNKT cells. In this study, intracellular Foxp3 staining was possible because chromatin-linked H2BeGFP fluorescence was not lost during cell permeabilization (21). Still, H2BeGFP intensities in fixed/permeabilized cells were found to be lower than in untreated cells. Thymic Treg precursors, defined as CD4+ SP Foxp3CD25+CD122+ (36), showed a 2-fold decline in H2BeGFP levels compared with semimature CD4+Foxp3 SP cells, but similar levels compared with mature CD4+Foxp3 SP (Fig. 7). This loss of H2BeGFP signal translates into one thymic cell division per cell after passing the CD4+CD8+ DP stage. Further maturation of thymic Tregs, characterized by Foxp3 induction, was not accompanied by a significant loss of H2BeGFP levels (Fig. 7). These observations are consistent with the idea that agonistic activation results in proliferation at Foxp3 Treg precursor stage. However, taking into account that H2BeGFP levels of CD4+ SP T cells were also divided in half during thymic maturation, this means that mature CD4+Foxp3 T cells and mature CD4+Foxp3+ Tregs leave the thymus with comparable H2BeGFP intensities, and thus with a similar proliferation history.

FIGURE 7.

Thymic Tregs proliferate at the immediate precursor stage. CD4+ SP CD24+CD69+ T cells were defined as semimature and CD24CD69 T cells as mature. Treg precursors were defined and gated as CD4+Foxp3CD25+CD122+ (upper panel). All samples were fixed and permeabilized with Foxp3 staining buffer (eBioscience) for intracellular Foxp3 staining and for comparison, respectively. Graph shows H2BeGFP MFI fold change of thymic semimature and mature CD4+Foxp3 SP T cells, Treg precursors, and Tregs from TcrdH2BeGFP mice compared with C57BL/6 controls. Combined data from two different experiments (n = 12) are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest. ***p < 0.001.

FIGURE 7.

Thymic Tregs proliferate at the immediate precursor stage. CD4+ SP CD24+CD69+ T cells were defined as semimature and CD24CD69 T cells as mature. Treg precursors were defined and gated as CD4+Foxp3CD25+CD122+ (upper panel). All samples were fixed and permeabilized with Foxp3 staining buffer (eBioscience) for intracellular Foxp3 staining and for comparison, respectively. Graph shows H2BeGFP MFI fold change of thymic semimature and mature CD4+Foxp3 SP T cells, Treg precursors, and Tregs from TcrdH2BeGFP mice compared with C57BL/6 controls. Combined data from two different experiments (n = 12) are shown. Statistical analysis was performed with one-way repeated-measures ANOVA with Tukey posttest. ***p < 0.001.

Close modal

Next, we analyzed proliferation of Tregs in the periphery in correlation to their anatomical location. In lymph nodes and spleen, Tregs showed lower H2BeGFP levels compared with CD4+Foxp3 T cells (Fig. 8). This suggests that, on average, peripheral Tregs in secondary lymphoid organs had proliferated more than CD4+Foxp3 T cells. This matches our previous findings that the peripheral Treg repertoire is very diverse, albeit less diverse than naive αβ T cells (25). In contrast, in small intestine and colon, Tregs and CD4+Foxp3 T cells completely lost their H2BeGFP, indicating that they proliferated more than two times after thymic exit (Fig. 8). It was proposed that different Treg subsets display distinct phenotypes and homeostasis in normal unmanipulated mice (37). In the steady-state, some Tregs remain quiescent and have a long life span, in the order of months, whereas the other Tregs are dividing extensively and express multiple activation markers. Therefore, we mapped Treg proliferation according to expression of activation markers. First, analogous to CD4+Foxp3 and CD8+ T cells, Tregs with a CD62LCD44+ phenotype had proliferated most (Fig. 9A), which is in line with previous findings (38). Next, we sought to investigate the proliferation history of Treg populations defined by Treg-specific markers. The transcription factor Helios (39) and the surface molecule Nrp1 (40, 41) were recently proposed to discriminate thymic-derived and peripherally induced Tregs. As revealed by H2BeGFP levels of the respective Treg populations in secondary lymphoid organs, Helios+ and Nrp1high Tregs had proliferated more than Helios or Nrp1low Tregs (Fig. 9B, 9C). An interesting exception was the mesenteric LN, where the proliferation history of Nrp1high and Nrp1low Tregs did not differ. In addition, activated FR4high Tregs (42) showed significantly lower H2BeGFP levels compared with FR4low and FR4intermediate Tregs (Fig. 9D). Taken together, expression of Helios, Nrp1, and FR4 correlated with postselection proliferation of Tregs, and thus likely indicated Treg activation.

FIGURE 8.

Peripheral Tregs proliferate more than CD4+Foxp3 T cells. Histograms show H2BeGFP level, and graph shows H2BeGFP MFI fold change of CD4+Foxp3 (gray) and CD4+Foxp3+ (black) T cells in peripheral lymph nodes and small intestine of TcrdH2BeGFP mice compared with C57BL/6 control (gray filled). Representative dot plot of CD4+ T cells of peripheral LNs from TcrdH2BeGFP mice. Graph shows quantification of fold change H2BeGFP MFI in CD4+Foxp3 (open) and CD4+Foxp3+ (filled) T cells derived from indicated organs compared with C57BL/6 control. Representative data from five different experiments (n = 19) are shown. Statistical analysis was performed with Student t test. ***p < 0.001.

FIGURE 8.

Peripheral Tregs proliferate more than CD4+Foxp3 T cells. Histograms show H2BeGFP level, and graph shows H2BeGFP MFI fold change of CD4+Foxp3 (gray) and CD4+Foxp3+ (black) T cells in peripheral lymph nodes and small intestine of TcrdH2BeGFP mice compared with C57BL/6 control (gray filled). Representative dot plot of CD4+ T cells of peripheral LNs from TcrdH2BeGFP mice. Graph shows quantification of fold change H2BeGFP MFI in CD4+Foxp3 (open) and CD4+Foxp3+ (filled) T cells derived from indicated organs compared with C57BL/6 control. Representative data from five different experiments (n = 19) are shown. Statistical analysis was performed with Student t test. ***p < 0.001.

Close modal
FIGURE 9.

Proliferation of peripheral Treg subsets. Different Treg subsets were gated according to expression of CD44, CD62L, Helios, Nrp1, and FR4. Graphs show H2BeGFP MFI fold change of CD62L+CD44 and CD62LCD44+ (A); Helios and Helios+ (B); Nrp1 and Nrp1+ (C); and FR4low, FR4intermediate, and FR4high (D) subsets of CD4+Foxp3+ Tregs from TcrdH2BeGFP mice compared with C57BL/6 control mice. (D) FR4 staining of cells from peripheral lymph nodes. Representative data from four [(A, C); n = 13], five [(B); n = 19], or two [(D); n = 8] different experiments are shown. Statistical analysis was performed with repeated-measures ANOVA and Tukey posttest. ***p < 0.001. nd, Not detectable.

FIGURE 9.

Proliferation of peripheral Treg subsets. Different Treg subsets were gated according to expression of CD44, CD62L, Helios, Nrp1, and FR4. Graphs show H2BeGFP MFI fold change of CD62L+CD44 and CD62LCD44+ (A); Helios and Helios+ (B); Nrp1 and Nrp1+ (C); and FR4low, FR4intermediate, and FR4high (D) subsets of CD4+Foxp3+ Tregs from TcrdH2BeGFP mice compared with C57BL/6 control mice. (D) FR4 staining of cells from peripheral lymph nodes. Representative data from four [(A, C); n = 13], five [(B); n = 19], or two [(D); n = 8] different experiments are shown. Statistical analysis was performed with repeated-measures ANOVA and Tukey posttest. ***p < 0.001. nd, Not detectable.

Close modal

In this study, TcrdH2BeGFP mice were used to directly track the proliferative history of αβ T cells by transient genetic pulse labeling during thymic development. This novel tool allowed us to revisit the contrasting postselection dynamics of mainstream αβ T cells, Foxp3+ Tregs, and iNKT cells. The latter, identified by α-galactosylceramide bound to CD1d (43), are representative innate T cells. Agonistic TCR selection of an invariant Vα14-Jα18 TCRα-chain (44), predominantly paired with Vβ8.2, Vβ7, and Vβ2, propagates intrathymic proliferation via a strong TCR signal (34). Recent data suggest that miR-181a/b-1 expression in DP thymocytes supports iNKT agonist selection and subsequent efficient expansion through intrathymic proliferation (45). In contrast, adaptive αβ T cells should maintain a highly diverse TCR repertoire to be prepared for recognition of any foreign Ag. Thus, it is conceivable that adaptive T cells should not proliferate unless they encounter their cognate ligands. Indeed, retention of H2BeGFP signal in naive peripheral CD4+ and CD8+ T cells indicated that these cells had, on average, undergone less than three cell divisions after positive selection and less than two divisions after thymic exit. Even in the highly inflammatory milieu of a fulminant primary immune response to MCMV infection, nonparticipating naive bystander T cells did not proliferate. This situation may be different in neonates, when the secondary lymphoid organs of mice are filling rapidly. The argument that little proliferation is a means of guaranteeing diversity of the TCR repertoire is valid for adult mice but may not be so for neonates. In adult mice, only strong lymphopenia after adoptive transfer into Rag−/−γc−/− recipients induced homeostatic proliferation; thereby CD8+ T cells appeared to be more responsive to homeostatic proliferation as described previously (46). Also, studies in RAG2p-GFP transgenic mice showed that, in particular, CD8+ RTEs undergo a process of postthymic maturation before they are incorporated into the pool of mature peripheral T cells (20). In RTEs, the CD4/CD8 ratio is higher compared with mature peripheral T cells, and adjustment of the CD4/CD8 ratio includes increased proliferation of CD8+ RTEs. In this study, we confirmed these results by demonstrating H2BeGFP loss during thymic exit solely for CD8+ and not for CD4+ RTEs. It was speculated that CD8+ thymocytes in RAG2p-GFP transgenic mice showed lower GFP levels because their DP to SP transition took longer time than in CD4+ thymocytes (19). In this study, we observed that decline of H2BeGFP reporter intensity was greater in CD8+ compared with CD4+ T cells. Because H2BeGFP is very stable, we therefore suggest slightly more proliferation during CD8+ DP-to-SP transition. Nevertheless, we observed proliferation in CD4+ as well as CD8+ SP thymocytes. Although H2BeGFP may be partially lost because of rebuilding of condense histone-rich chromosomal regions during thymocyte differentiation, one might speculate that one to two thymic cell divisions produce a “backup” of the naive TCR repertoire.

It is currently controversial whether self-Ag recognition by thymic Treg precursors equals agonist selection and to what extent this Ag encounter induces proliferation (9, 10, 12, 34, 35). Hogquist’s group (34) showed with Nur77GFP reporter mice that Tregs, upon thymic selection, perceived a stronger TCR stimulation compared with CD4+Foxp3 T cells. Unexpectedly, thymic Tregs in TcrdH2BeGFP mice displayed reporter fluorescence levels similar to mature CD4+Foxp3 SP cells. This supports the idea that thymus re-entry or long-term thymic retention, as described for agonist-selected iNKT (47) or γδ NKT cells (48), does not chiefly contribute to the thymic Treg pool.

Thymus-derived Tregs make up the largest part of the peripheral Treg compartment. However, Tregs can also be induced from CD4+Foxp3 peripheral T cells. Up to now, no solid marker to discriminate thymus-derived and peripherally induced Tregs was found. Recently, Nrp-1 was proposed to be expressed at high levels on thymus-derived opposed to peripherally induced Tregs (40, 41). In this study, we found lower H2BeGFP levels in Nrp1high compared with Nrp1low Tregs. Furthermore, Helios, an Ikaros transcription factor family member, was proposed to be expressed by thymus-derived but not peripherally induced Tregs (39, 49, 50). However, Helios might be rather a marker for T cell activation and proliferation (51). In line with the latter notion, we could show that Helios+ Tregs are significantly more proliferated than their Helios counterparts. Together, assuming that lack of Nrp-1 and Helios expression identifies peripherally induced Tregs, their Foxp3+ induction would not be accompanied by proliferation.

In conclusion, these studies further illustrate that the thymus maintains two fundamentally different production lines: innate and adaptive T cells. On one hand, agonist-selected iNKT cells display a limited TCR repertoire, and they are efficiently produced via intrathymic proliferation. On the other hand, adaptive αβ T cells hardly proliferate during thymic development and maintain a highly diverse TCR repertoire. Thymic Tregs, although possibly activated by TCR agonist selection, do not proliferate more than CD4+Foxp3 thymocytes and also sustain a highly diverse TCR repertoire. In future work, being able to track the proliferative history of αβ T cells at the individual cell level in TcrdH2BeGFP mice should be useful to study homeostatic proliferation and loss of clonal diversity in aged mice.

We thank Andreas Krueger for discussions and carefully reading the manuscript. We thank Sylvie Guerder, Scott Robins, Mathias Herberg, Manuel Winter, André Bleich, Anna Smoczek, Jens Pommerencke, Marcin Łyszkiewicz, and Natalia Ziętara for help with this work. We acknowledge the assistance of the Cell Sorting Core Facility of the Hannover Medical School.

This work was supported by the Deutsche Forschungsgemeinschaft (Grants SFB621-A14, PR727/4-1, PR727/5-1 to I.P.), the Deutsche José Carreras Leukämie-Stiftung (Project Grant R 12/29 to I.P.), and the Deutsch-Französische Hochschule/Université Franco-Allemande (Grant DFH-UFA G2RFA 104-07-II to I.P. and B.M.). L.F. is a scholar of the Hannover Biomedical Research School.

The online version of this article contains supplemental material.

Abbreviations used in this article:

DN

double-negative

DP

double-positive

H2BeGFP

histone 2B–enhanced GFP

iNKT

invariant NKT

MCMV

murine CMV

MFI

median H2BeGFP fluorescence intensity

Nrp1

Neuropilin-1

RTE

recent thymic emigrant

Treg

regulatory T cell.

1
Laurent
J.
,
Bosco
N.
,
Marche
P. N.
,
Ceredig
R.
.
2004
.
New insights into the proliferation and differentiation of early mouse thymocytes.
Int. Immunol.
16
:
1069
1080
.
2
Zlotoff
D. A.
,
Bhandoola
A.
.
2011
.
Hematopoietic progenitor migration to the adult thymus.
Ann. N. Y. Acad. Sci.
1217
:
122
138
.
3
Dudley
E. C.
,
Petrie
H. T.
,
Shah
L. M.
,
Owen
M. J.
,
Hayday
A. C.
.
1994
.
T cell receptor beta chain gene rearrangement and selection during thymocyte development in adult mice.
Immunity
1
:
83
93
.
4
Kreslavsky
T.
,
Gleimer
M.
,
Miyazaki
M.
,
Choi
Y.
,
Gagnon
E.
,
Murre
C.
,
Sicinski
P.
,
von Boehmer
H.
.
2012
.
β-Selection-induced proliferation is required for αβ T cell differentiation.
Immunity
37
:
840
853
.
5
Le Campion
A.
,
Vasseur
F.
,
Pénit
C.
.
2000
.
Regulation and kinetics of premigrant thymocyte expansion.
Eur. J. Immunol.
30
:
738
746
.
6
Pénit
C.
,
Vasseur
F.
.
1997
.
Expansion of mature thymocyte subsets before emigration to the periphery.
J. Immunol.
159
:
4848
4856
.
7
Bendelac
A.
,
Savage
P. B.
,
Teyton
L.
.
2007
.
The biology of NKT cells.
Annu. Rev. Immunol.
25
:
297
336
.
8
Benlagha
K.
,
Wei
D. G.
,
Veiga
J.
,
Teyton
L.
,
Bendelac
A.
.
2005
.
Characterization of the early stages of thymic NKT cell development.
J. Exp. Med.
202
:
485
492
.
9
Bautista
J. L.
,
Lio
C. W.
,
Lathrop
S. K.
,
Forbush
K.
,
Liang
Y.
,
Luo
J.
,
Rudensky
A. Y.
,
Hsieh
C. S.
.
2009
.
Intraclonal competition limits the fate determination of regulatory T cells in the thymus.
Nat. Immunol.
10
:
610
617
.
10
Leung
M. W.
,
Shen
S.
,
Lafaille
J. J.
.
2009
.
TCR-dependent differentiation of thymic Foxp3+ cells is limited to small clonal sizes.
J. Exp. Med.
206
:
2121
2130
.
11
Ribot
J.
,
Romagnoli
P.
,
van Meerwijk
J. P.
.
2006
.
Agonist ligands expressed by thymic epithelium enhance positive selection of regulatory T lymphocytes from precursors with a normally diverse TCR repertoire.
J. Immunol.
177
:
1101
1107
.
12
Romagnoli
P.
,
Dooley
J.
,
Enault
G.
,
Vicente
R.
,
Malissen
B.
,
Liston
A.
,
van Meerwijk
J. P.
.
2012
.
The thymic niche does not limit development of the naturally diverse population of mouse regulatory T lymphocytes.
J. Immunol.
189
:
3831
3837
.
13
Taylor
J. H.
,
Woods
P. S.
,
Hughes
W. L.
.
1957
.
The organization and duplication of chromosomes as revealed by autoradiographic studies using tritium-labeled thymidinee.
Proc. Natl. Acad. Sci. USA
43
:
122
128
.
14
Gratzner
H. G.
1982
.
Monoclonal antibody to 5-bromo- and 5-iododeoxyuridine: a new reagent for detection of DNA replication.
Science
218
:
474
475
.
15
Lyons
A. B.
,
Parish
C. R.
.
1994
.
Determination of lymphocyte division by flow cytometry.
J. Immunol. Methods
171
:
131
137
.
16
Gerdes
J.
,
Lemke
H.
,
Baisch
H.
,
Wacker
H. H.
,
Schwab
U.
,
Stein
H.
.
1984
.
Cell cycle analysis of a cell proliferation-associated human nuclear antigen defined by the monoclonal antibody Ki-67.
J. Immunol.
133
:
1710
1715
.
17
Ernst
B.
,
Surh
C. D.
,
Sprent
J.
.
1995
.
Thymic selection and cell division.
J. Exp. Med.
182
:
961
971
.
18
Yu
W.
,
Nagaoka
H.
,
Jankovic
M.
,
Misulovin
Z.
,
Suh
H.
,
Rolink
A.
,
Melchers
F.
,
Meffre
E.
,
Nussenzweig
M. C.
.
1999
.
Continued RAG expression in late stages of B cell development and no apparent re-induction after immunization.
Nature
400
:
682
687
.
19
McCaughtry
T. M.
,
Wilken
M. S.
,
Hogquist
K. A.
.
2007
.
Thymic emigration revisited.
J. Exp. Med.
204
:
2513
2520
.
20
Boursalian
T. E.
,
Golob
J.
,
Soper
D. M.
,
Cooper
C. J.
,
Fink
P. J.
.
2004
.
Continued maturation of thymic emigrants in the periphery.
Nat. Immunol.
5
:
418
425
.
21
Houston
E. G.
 Jr.
,
Fink
P. J.
.
2009
.
MHC drives TCR repertoire shaping, but not maturation, in recent thymic emigrants.
J. Immunol.
183
:
7244
7249
.
22
Prinz
I.
,
Sansoni
A.
,
Kissenpfennig
A.
,
Ardouin
L.
,
Malissen
M.
,
Malissen
B.
.
2006
.
Visualization of the earliest steps of gammadelta T cell development in the adult thymus.
Nat. Immunol.
7
:
995
1003
.
23
Tumbar
T.
,
Guasch
G.
,
Greco
V.
,
Blanpain
C.
,
Lowry
W. E.
,
Rendl
M.
,
Fuchs
E.
.
2004
.
Defining the epithelial stem cell niche in skin.
Science
303
:
359
363
.
24
Lathrop
S. K.
,
Santacruz
N. A.
,
Pham
D.
,
Luo
J.
,
Hsieh
C. S.
.
2008
.
Antigen-specific peripheral shaping of the natural regulatory T cell population.
J. Exp. Med.
205
:
3105
3117
.
25
Föhse
L.
,
Suffner
J.
,
Suhre
K.
,
Wahl
B.
,
Lindner
C.
,
Lee
C. W.
,
Schmitz
S.
,
Haas
J. D.
,
Lamprecht
S.
,
Koenecke
C.
, et al
.
2011
.
High TCR diversity ensures optimal function and homeostasis of Foxp3+ regulatory T cells.
Eur. J. Immunol.
41
:
3101
3113
.
26
Prinz
I.
2011
.
Dynamics of the interaction of γδ T cells with their neighbors in vivo.
Cell. Mol. Life Sci.
68
:
2391
2398
.
27
Legendre
V.
,
Boyer
C.
,
Guerder
S.
,
Arnold
B.
,
Hämmerling
G.
,
Schmitt-Verhulst
A. M.
.
1999
.
Selection of phenotypically distinct NK1.1+ T cells upon antigen expression in the thymus or in the liver.
Eur. J. Immunol.
29
:
2330
2343
.
28
Robbins
S. H.
,
Tessmer
M. S.
,
Mikayama
T.
,
Brossay
L.
.
2004
.
Expansion and contraction of the NK cell compartment in response to murine cytomegalovirus infection.
J. Immunol.
173
:
259
266
.
29
Orange
J. S.
,
Wang
B.
,
Terhorst
C.
,
Biron
C. A.
.
1995
.
Requirement for natural killer cell-produced interferon gamma in defense against murine cytomegalovirus infection and enhancement of this defense pathway by interleukin 12 administration.
J. Exp. Med.
182
:
1045
1056
.
30
Lindsten
T.
,
Fowlkes
B. J.
,
Samelson
L. E.
,
Davis
M. M.
,
Chien
Y. H.
.
1987
.
Transient rearrangements of the T cell antigen receptor alpha locus in early thymocytes.
J. Exp. Med.
166
:
761
775
.
31
Koszinowski
U. H.
1991
.
Molecular aspects of immune recognition of cytomegalovirus.
Transplant. Proc.
23
(
3
Suppl. 3
):
70
73, discussion 74
.
32
Krmpotic
A.
,
Bubic
I.
,
Polic
B.
,
Lucin
P.
,
Jonjic
S.
.
2003
.
Pathogenesis of murine cytomegalovirus infection.
Microbes Infect.
5
:
1263
1277
.
33
Benlagha
K.
,
Kyin
T.
,
Beavis
A.
,
Teyton
L.
,
Bendelac
A.
.
2002
.
A thymic precursor to the NK T cell lineage.
Science
296
:
553
555
.
34
Moran
A. E.
,
Holzapfel
K. L.
,
Xing
Y.
,
Cunningham
N. R.
,
Maltzman
J. S.
,
Punt
J.
,
Hogquist
K. A.
.
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
.
35
Jordan
M. S.
,
Boesteanu
A.
,
Reed
A. J.
,
Petrone
A. L.
,
Holenbeck
A. E.
,
Lerman
M. A.
,
Naji
A.
,
Caton
A. J.
.
2001
.
Thymic selection of CD4+CD25+ regulatory T cells induced by an agonist self-peptide.
Nat. Immunol.
2
:
301
306
.
36
Lio
C. W.
,
Hsieh
C. S.
.
2008
.
A two-step process for thymic regulatory T cell development.
Immunity
28
:
100
111
.
37
Fisson
S.
,
Darrasse-Jèze
G.
,
Litvinova
E.
,
Septier
F.
,
Klatzmann
D.
,
Liblau
R.
,
Salomon
B. L.
.
2003
.
Continuous activation of autoreactive CD4+ CD25+ regulatory T cells in the steady state.
J. Exp. Med.
198
:
737
746
.
38
Min
B.
,
Thornton
A.
,
Caucheteux
S. M.
,
Younes
S. A.
,
Oh
K.
,
Hu-Li
J.
,
Paul
W. E.
.
2007
.
Gut flora antigens are not important in the maintenance of regulatory T cell heterogeneity and homeostasis.
Eur. J. Immunol.
37
:
1916
1923
.
39
Thornton
A. M.
,
Korty
P. E.
,
Tran
D. Q.
,
Wohlfert
E. A.
,
Murray
P. E.
,
Belkaid
Y.
,
Shevach
E. M.
.
2010
.
Expression of Helios, an Ikaros transcription factor family member, differentiates thymic-derived from peripherally induced Foxp3+ T regulatory cells.
J. Immunol.
184
:
3433
3441
.
40
Yadav, M., C. Louvet, D. Davini, J. M. Gardner, M. Martinez-Llordella, S. Bailey-Bucktrout, B. A. Anthony, F. M. Sverdrup, R. Head, D. J. Kuster, et al. 2012. Neuropilin-1 distinguishes natural and inducible regulatory T cells among regulatory T cell subsets in vivo. J. Exp. Med. 209: 1713–1722, S1–S19.
41
Weiss, J. M., A. M. Bilate, M. Gobert, Y. Ding, M. A. Curotto de Lafaille, C. N. Parkhurst, H. Xiong, J. Dolpady, A. B. Frey, M. G. Ruocco, et al. 2012. Neuropilin 1 is expressed on thymus-derived natural regulatory T cells, but not mucosa-generated induced Foxp3+ T reg cells. J. Exp. Med. 209: 1723–1742, S1.
42
Yamaguchi
T.
,
Hirota
K.
,
Nagahama
K.
,
Ohkawa
K.
,
Takahashi
T.
,
Nomura
T.
,
Sakaguchi
S.
.
2007
.
Control of immune responses by antigen-specific regulatory T cells expressing the folate receptor.
Immunity
27
:
145
159
.
43
Bendelac
A.
,
Lantz
O.
,
Quimby
M. E.
,
Yewdell
J. W.
,
Bennink
J. R.
,
Brutkiewicz
R. R.
.
1995
.
CD1 recognition by mouse NK1+ T lymphocytes.
Science
268
:
863
865
.
44
Lantz
O.
,
Bendelac
A.
.
1994
.
An invariant T cell receptor alpha chain is used by a unique subset of major histocompatibility complex class I-specific CD4+ and CD4-8- T cells in mice and humans.
J. Exp. Med.
180
:
1097
1106
.
45
Ziętara
N.
,
Łyszkiewicz
M.
,
Witzlau
K.
,
Naumann
R.
,
Hurwitz
R.
,
Langemeier
J.
,
Bohne
J.
,
Sandrock
I.
,
Ballmaier
M.
,
Weiss
S.
, et al
.
2013
.
Critical role for miR-181a/b-1 in agonist selection of invariant natural killer T cells.
Proc. Natl. Acad. Sci. USA
110
:
7407
7412
.
46
Berzins
S. P.
,
Boyd
R. L.
,
Miller
J. F.
.
1998
.
The role of the thymus and recent thymic migrants in the maintenance of the adult peripheral lymphocyte pool.
J. Exp. Med.
187
:
1839
1848
.
47
Berzins
S. P.
,
McNab
F. W.
,
Jones
C. M.
,
Smyth
M. J.
,
Godfrey
D. I.
.
2006
.
Long-term retention of mature NK1.1+ NKT cells in the thymus.
J. Immunol.
176
:
4059
4065
.
48
Grigoriadou
K.
,
Boucontet
L.
,
Pereira
P.
.
2003
.
Most IL-4-producing gamma delta thymocytes of adult mice originate from fetal precursors.
J. Immunol.
171
:
2413
2420
.
49
Gottschalk
R. A.
,
Corse
E.
,
Allison
J. P.
.
2012
.
Expression of Helios in peripherally induced Foxp3+ regulatory T cells.
J. Immunol.
188
:
976
980
.
50
Verhagen
J.
,
Wraith
D. C.
.
2010
.
Comment on “Expression of Helios, an Ikaros transcription factor family member, differentiates thymic-derived from peripherally induced Foxp3+ T regulatory cells”.
J. Immunol.
185
:
7129
,
author reply 7130
.
51
Akimova
T.
,
Beier
U. H.
,
Wang
L.
,
Levine
M. H.
,
Hancock
W. W.
.
2011
.
Helios expression is a marker of T cell activation and proliferation.
PLoS ONE
6
:
e24226
.

The authors have no financial conflicts of interest.