common γ chain
Marilyn TCR Tg mice
: protein kinase B
Cytokines signaling through receptors sharing the common γ chain (γc), including IL-2, IL-4, IL-7, IL-9, IL-15 and IL-21, are critical for the generation and peripheral homeostasis of B, T and NK cells. To identify unique or redundant roles for γc cytokines in naive CD4+ T cells, we compared monoclonal populations of CD4+ T cells from TCR-Tg mice that were γ, γ, CD127–/– or CD122–/–. We found that γ naive CD4+ T cells failed to accumulate in the peripheral lymphoid organs and the few remaining cells were characterized by small size, decreased expression of MHC class I and enhanced apoptosis. By over-expressing human Bcl-2, peripheral naive CD4+ T cells that lack γc could be rescued. Bcl-2+ γ CD4+ T cells demonstrated enhanced survival characteristics in vivo and in vitro, and could proliferate normally in vitro in response to antigen. Nevertheless, Bcl-2+ γ CD4+ T cells remained small in size, and this phenotype was not corrected by enforced expression of an activated protein kinase B. We conclude that γc cytokines (primarily but not exclusively IL-7) provide Bcl-2-dependent as well as Bcl-2-independent signals to maintain the phenotype and homeostasis of the peripheral naive CD4+ T cell pool.
See accompanying commentary: http://dx.doi.org/10.1002/eji.200737721
Cytokines play essential roles in lymphocyte homeostasis. During lymphoid development, distinct cytokines have been identified that promote the survival and proliferation of B, T and NK cell precursors in the bone marrow and in the thymus. For example, IL-7 and stem cell factor are crucial for early thymocyte differentiation, IL-7 and fetal liver kinase-2 ligand drive B cell development, and IL-15 is essential for the generation of immature NK cells 1–5. These observations indicate that cytokines have specific (i.e. non-redundant) roles during early lymphocyte development.
Cytokines that signal through the common γ chain (γc) include IL-2, IL-4, IL-7, IL-9, IL-15 and IL-21. Upon ligand binding, γc-containing receptors transmit intracellular signals through the JAK1 and JAK3 tyrosine kinases to activate primarily the transcription factors STAT3 and STAT5 6. Several γc-dependent receptors also signal through the PI3K, protein kinase B (PKB), MAPK and src family kinase pathways 7 and result in additional transcription factor activation. These different activation pathways converge in the nucleus where they modify gene expression profiles, although the impact of distinct transcription factors in this process is still unclear. Considering the complexity of γc signaling, it is not surprising that the transcriptional changes elicited by distinct γc cytokines include both unique and shared targets 8.
Downstream targets of γc signaling in lymphocytes include proteins involved in cell survival and apoptosis, such as members of the Bcl-2 family. This protein family includes both pro- and anti-apoptotic members; the latter inhibit the former by specific heterodimeric interactions 9. Bcl-2 is a major target in early thymocyte development, especially during the IL-7-dependent ‘double-negative (DN)2 stage 10. In the absence of IL-7/γc, DN2 cells have decreased Bcl-2 levels, are prone to apoptosis and fail to proliferate normally 11.
Nevertheless, Bcl-2 is not the sole target of IL-7 signals at this stage, since over-expression of a Bcl-2 Tg in γ or CD127 (IL-7Rα)–/– mice does not correct the transitional DN2 block 12, although it improves the overall T cell homeostasis, especially in the periphery of IL-7 signaling-deficient animals 13, 14. One possibility is that other Bcl-2 family members are also regulated by IL-7 in early thymocytes, including the pro-apoptotic Bax and/or Bad proteins 15. Along these lines, mice made deficient in both CD127 and Bax demonstrated increased thymic cellularity compared with CD127– mice 16. Finally, the anti-apoptotic protein Mcl-1 is also strongly up-regulated by IL-7 in thymocytes and peripheral T cells 17. Mcl-1 acts independently of Bcl-2, and could explain the inability of enforced Bcl-2 expression to completely rescue the DN block in γ mice, or to allow recovery of other IL-7-dependent lymphocyte populations (i.e. γδ T cells and B cells) in this context 18.
Once thymocytes complete the selection process and have fully differentiated as mature CD4 or CD8 single-positive (SP) thymocytes, they exit the thymus and re-circulate through the secondary lymphoid organs, including the lymph nodes and spleen. In the periphery, T cells require signals for their maintenance, which can be conveyed by TCR engagement, cell surface co-stimulatory receptors or soluble factors including cytokines. Of the latter, γc-dependent cytokines play critical roles for the homeostasis of naive αβ T cells, γδ T cells, NK and NKT cells. Adoptive transfer studies have shown that IL-7 supports peripheral survival and homeostatic proliferation of ‘adaptive’ CD4+ and CD8+ αβ T cells, while IL-15 is essential for survival of ‘innate’ NK, NKT and γδ T cells (reviewed in 19).
The downstream effectors of IL-7 and IL-15 in this context remain to be fully defined, although these cytokines can regulate the expression of Bcl-2 family proteins 10, 20. With respect to the role of IL-7 in naive T cell homeostasis, culture of human or mouse naive T cells in vitro with γc cytokines up-regulates Bcl-2 expression 21 and over-expression of Bcl-2 increases peripheral αβ T cell numbers in CD127–/– mice 13, 14. In contrast, Tg over-expression of Bcl-2 in γ T cells showed a limited effect on numbers of peripheral naive T cells 18. This observation might indicate that γc-dependent cytokines other than IL-7 are involved in the homeostasis of naive T cells. Alternatively, these results might have been influenced by the propensity of γc-deficient mice to develop an inflammatory syndrome (secondary to generalized T cell activation resulting from an absence of regulatory T cells 22). Our knowledge of the biological roles played by γc cytokines in the maintenance of naive T cells remains incomplete.
In this report, we utilize Tg mice bearing the male antigen-specific Marilyn (Ml) TCR to assess the role for γc cytokines in the biology of naive CD4+ T cells. In order to limit spurious TCR specificities resulting from associations of the Tg TCR β chain with endogenously rearranged TCR α polypeptides, all mice were RAG–/–, and thereby harbored ‘monoclonal’ T cell populations. By comparing γ, γ, CD127–/– and CD122–/– Ml female mice, we define a series of γc-dependent phenotypes in peripheral naive CD4+ T cells. By over-expressing Bcl-2 or a constitutively activated form of PKB in γ and γ Ml CD4+ T cells, we have further characterized the signaling pathways required for γc-dependent naive CD4+ T cell homeostasis. Our results demonstrate that γc cytokines (primarily IL-7) provide Bcl-2-dependent as well as Bcl-2-indpendent signals to peripheral naive CD4+ T cells.
Phenotype of γ naive CD4+ T cells
The TCR-Tg model system Ml on the RAG2–/– background generates monoclonal populations of CD4+ αβ T cells reactive with the male antigen Dby 23. Peripheral CD4+ T cells in Ml female Tg mice have a naive phenotype (CD44lo CD62Lhi), are non-cycling and demonstrate limited homeostatic proliferation after transfer to lymphopenic recipients 24, consistent with their limited cross-reactivity to environmental and/or self antigens. In the absence of γc, CD4 SP Ml T cells are efficiently selected in the thymus, but fail to accumulate in the peripheral lymphoid organs 23. These peripheral γ Ml CD4+ T cells are reduced about 200-fold in comparison to their WT counterparts, although they demonstrate the expected naive CD44lo CD62Lhi phenotype. Moreover, γ Ml CD4+ T cells in the spleen had markedly reduced Bcl-2 levels, suggesting a survival defect 23.
In order to gain more insights into the unique or redundant roles for γc cytokines in the homeostasis of naive CD4+ T cells, we further characterized the cell surface phenotype of splenic CD4+ Ml T cells that developed in the absence of γc (Fig. 1A–D). We found that splenic γ and γ Ml CD4+ T cells expressed similar levels of CD2, Vβ6, CD4, CD44 and CD62L and were negative for CD69 (Fig. 1A, B and data not shown). Thus, unlike non-TCR-Tg γ mice, T cells from Ml γ female mice had a naive phenotype.
Concerning cytokine receptors, CD127 expression was normal in the absence of γc, while naive Ml CD4+ T cells failed to express CD25 or CD122 (Fig. 1C). Consistent with their strongly reduced Bcl-2 levels 23, we found that γ Ml CD4+ T cells had an abnormal mitochondrial membrane potential as assessed by staining with DIOC6 and were smaller in size than their WT counterparts (Fig. 1D). These results provide an in vivo confirmation of the previous report demonstrating that IL-7 can maintain cell size and viability of naive T cells in vitro25. We further found that γ Ml CD4+ T cells expressed lower levels of MHC class I molecules (Fig. 1D). These results suggest that γc signaling in vivo affects the naive T cell phenotype at multiple levels.
We next compared the phenotype of the peripheral cells to the mature CD4 SP thymocytes in γ or γ Ml female mice to assess whether the phenotypes observed were a simple consequence of γc deficiency, or whether they resulted from a selective pressure that was imposed in the periphery (Fig. 2A–C). Compared to their γ counterparts, CD4 SP thymocytes from γ Ml female mice bore identical levels of CD4, Vβ6 and CD127 (Fig. 2B). In contrast, γ CD4 SP thymocytes were smaller in size compared to WT cells and expressed lower levels of H-2Db (Fig. 2C). These results indicate that γc cytokines already begin to play a homeostatic role at the mature CD4 SP thymocyte stage.
Enforced Bcl-2 expression restores peripheral γ naive CD4+ T cell numbers
We next determined whether the phenotypes associated with γc deficiency could be corrected by enforced expression of Bcl-2. Previous studies had shown that Bcl-2 could substantially improve αβ T cell development in mice deficient in IL-7 signaling 13, 14, although it was unclear whether this corrected a defect in the naive cell compartment or improved survival of activated memory T cells that are over-represented in CD127–/– mice 26. To target Bcl-2 to the naive CD4+ T cell compartment, we generated ‘monoclonal’ Ml female mice that over-expressed human Bcl-2 in T cells (using the Eμ-2-25 line; 27) and were either γ or γ.
Tg expression of Bcl-2 had little effect on the absolute number of thymocytes in γ or γ Ml female mice (Fig. 3A), consistent with previous reports on the inability of this Bcl-2 Tg to correct the DN2 block in early thymocyte development in the absence of γc12. Nevertheless, Bcl-2 dramatically corrected the peripheral T cell defect in these mice (increase of 50-fold), allowing for a near normal number of splenic naive CD4+ T cells despite the absence of γc (Fig. 3A, B). This result suggested that poor survival of γ Ml CD4+ T cells was a major cause for their reduced cell numbers in the spleen (but not in the thymus). Accordingly, Bcl-2+ γ Ml CD4+ T cells had a normal mitochondrial membrane potential (Fig. 3C). Cell surface expression of CD4, Vβ6, CD44, CD62L and CD127 was not modified by Bcl-2 over-expression (Fig. 3C and data not shown).
When cultured in vitro in the absence of exogenous γc cytokines, Bcl-2+ Ml CD4+ T cells (either γ or γ) demonstrated enhanced survival compared to non-Bcl-2+ γ Ml CD4+ T cells (data not shown). Interestingly, despite their enhanced survival characteristics in vivo and in vitro, Bcl-2+ γ Ml CD4+ T cells remained small in size. Moreover, these cells continued to have reduced expression of MHC class I (Fig. 3D) and showed a generalized reduction in ribosomal protein transcripts (Supporting Information Table 1). These results confirm and extend previous in vitro studies on the role of IL-7 and Bcl-2 in naive T cell homeostasis 21, 25 and demonstrate that Bcl-2-independent pathways are triggered in vivo by γc cytokines in naive T cells to control cell size and expression of some cell surface proteins.
The ability of exogenous Bcl-2 to rescue a substantial population of γ naive CD4+ T cells provided the opportunity to formally assess the role for γc cytokines in antigen-induced T cell proliferation in vitro. IL-2 was initially characterized as ‘T cell growth factor’ and numerous studies have amply demonstrated the capacity for IL-2 to promote T cell proliferation in culture (reviewed in 28). We labeled splenocytes from Bcl-2+ Ml γ or γ female mice with CFSE and cultured the cells in the presence of antigen-presenting cells from female or male CD3ϵ– mice. We observed robust proliferation of the Bcl-2+ Ml CD4+ T cells after antigen stimulation and after several rounds of division, activated T cells up-regulated their expression of several cell surface markers (including CD25, CD44 and CD69) and down-regulated CD62L (Fig. 4A–D and data not shown), while increasing their cell size.
Although a similar profile of antigen-dependent activation was observed in the absence of γc, the kinetics showed that γ cells had a slight delay at day 4 post-stimulation, which was not apparent at day 6 (Fig. 4D). These results indicate that γc cytokines were redundant for in vitro proliferation provided that their survival was maintained by Bcl-2. Control Ml γ CD4+ T cells lacking Bcl-2 showed a similar response (data not shown), suggesting that Bcl-2 over-expression did not dramatically alter the capacity of Ml CD4+ T cells to undergo a normal differentiation program in response to male antigen-presenting cells. Finally, despite their reduced MHC class I expression, Ml γ CD4+ T cells were not selectively targeted for elimination by NK cells, despite the presence of the latter in the cultures (data not shown).
Reduced cell size in γ CD4+ T cells expressing activated PKB
Cytokines appear to regulate naive T cell size in the ‘resting’ state and can provoke a ‘blastogenic’ response characterized by increased cell size and enhanced nutrient uptake (reviewed in 29, 30). Since activation of the intracellular kinase PKB (also known as Akt) in response to growth factors can trigger increases in cell size, metabolism and survival in several cellular systems 29, 30, it has been proposed that IL-7 may act through PKB to mediate its homeostatic effects. Previous studies have shown that an activated PKB Tg could enhance T cell survival in vivo31, 32, although these Tg mice developed autoimmunity and lymphoma, suggesting PKB-mediated transformation. Moreover, conditional deletion of phosphatase and tension homologue deleted on chromosome 10 that counteracts PKB, can bypass the dependence on IL-7 and pre-TCR signals in developing thymocytes 33, although again these mice developed T cell lymphomas. In both these cases, the PKB expression in the peripheral compartment was not restricted to naive T cells. We therefore asked whether a PKB Tg expressed in developing and mature T cells 34 would modify the γc deficiency phenotypes that we had observed in naive Ml CD4+ T cells.
We generated monoclonal Ml γ female mice that were also Tg for Bcl-2 and/or activated PKB. Unlike Bcl-2, the PKB Tg only slightly increased numbers (about threefold) of naive CD4+ T cells in the spleen (Fig. 5A, B, G). Moreover, co-expression of Bcl-2 and PKB Tg did not change the phenotype of naive γ CD4+ T cells beyond that already observed after enforced Bcl-2 expression (Fig. 5C, D). Ml γ CD4+ T cells expressing the PKB Tg remained small in size compared to their γ counterparts (Fig. 5E, F). Thus, PKB expression in naive T cells is not sufficient to correct the phenotype of γc deficiency. In contrast to pharmacological inhibitors that block the PI3K/PKB activation and result in decreased cell size 25, 35, the absence of PKB activation does not appear responsible for the small cell size observed in Ml γ CD4+ T cells in vivo.
Involvement of CD127 but not CD122 in the homeostasis of peripheral CD4+ T cells
In order to identify the γc-dependent cytokines that are responsible for CD4+ T cell homeostasis, we generated monoclonal Ml female mice deficient in either CD122 or CD127 and then compared the development of naive CD4+ T cells in these mice with that of their WT counterparts. In the thymus, the absence of CD122 had no discernable effect on thymocyte differentiation, and absolute numbers of total thymocytes and CD4 SP T cells were normal (Fig. 6A). CD4 SP thymocyte size was unaltered (data not shown). In contrast, CD127 deficiency dramatically decreased total thymocyte cell numbers, although positive selection of CD4 SP thymocytes proceeded efficiently (Fig. 6C), with total numbers of CD4 SP thymocytes reaching near normal values (Fig. 6E), similar to what reported previously in the absence of γc23. These results confirm earlier reports demonstrating that IL-7 (and potentially TSLP) but neither IL-2 nor IL-15 are involved in the generation of the SP CD4 thymocyte pool 19.
Deficiency in CD122 had no obvious effect on the peripheral homeostasis of Ml CD4+ T cells (Fig. 6B). Absolute numbers of splenic CD122–/– Ml CD4+ T cells, their size, mitochondrial membrane potential and cell surface phenotype were similar to WT Ml CD4+ T cells (Fig. 6E and data not shown). In contrast, CD127 deficiency resulted in a marked reduction of Ml CD4+ T cells in the spleen; residual CD4+ T cells were small in size and had reduced DIO6 staining (Fig. 6F). Curiously, cell surface H-2Db staining was only slightly decreased on splenic CD4+ T cells from CD127–/– Ml female mice (Fig. 6F), suggesting that γc cytokines other than IL-7 may be important for this phenotype. Moreover, careful comparison of splenic CD4+ T cell numbers in γ and CD127–/– Ml female mice showed that the former were significantly reduced compared to the latter (Figs. 3A, 6E; p<0.05), consistent with the notion that naive T cell homeostasis depends on γc cytokines beyond IL-7.
Previous studies had shown that Bcl-2 could not replace the requirement for IL-7 in the homeostatic expansion of peripheral CD8+ T cells 36. We assessed the in vivo role for IL-7 and Bcl-2 expression for the survival and homeostatic expansion of naive Ml CD4+ T cells using an adoptive transfer approach. WT or Bcl-2+ Ml CD4+ splenic T cells were labeled with CFSE and transferred to alymphoid hosts that were IL-7+ or IL-7–. After 2 wk, the recipients were sacrificed and Ml CD4+ T cells were enumerated and their CFSE profiles analyzed.
Consistent with earlier reports 24, we found that Ml CD4+ T cells underwent two to three rounds of division after transfer to alymphoid recipients, while few T cells were recovered after transfer to IL-7-deficient hosts (Fig. 7). Bcl-2+ Ml CD4+ T cells showed a similar pattern of ‘homeostatic’ proliferation after transfer to alymphoid hosts as their Bcl-2 Tg– counterparts. In contrast, Bcl-2+ Ml CD4+ T cells survived well in vivo in the absence of IL-7, although they remained undivided (Fig. 7). These observations confirm the role for IL-7 in the survival of peripheral T cells 36 and further demonstrate that IL-7 is required in a Bcl-2-independent fashion for the homeostatic proliferation of naive CD4+ T cells.
The roles played by γc cytokines in the homeostasis of naive T cells are only partly defined. The use of γ mice to address this question was hampered by the spontaneous T cell activation that results as polyclonal repertoires with the potential to react with environmental antigens expand in the absence of regulatory T cells 37, 38. We therefore utilized a model system based on monoclonal TCR-Tg CD4+ T cells (MI mice) with specificity for a male antigen peptide presented by MHC class II. In female Ml mice, these cells are positively selected, but once exported to the peripheral lymphoid organs show little environmental cross-reactivity. This particular characteristic of Ml-Tg mice was essential, since it allowed us then to generate and analyze γ Ml CD4+ T cells that maintained a naive phenotype and to assess their properties ex vivo.
Using this approach, we were able to show that γ naïve T cells exhibit decreased survival parameters (mitochondrial membrane potential, reduced expression of Bcl-2), were small in size and had reduced expression of MHC class I molecules. These phenotypic characteristics did not appear to result from a selection process imposed by the peripheral naive T cell niche, since they were also observed in the mature CD4+ thymocyte compartment. By restoring Bcl-2 expression to developing thymocytes and mature γ naive T cells, we could show that some of the phenotypic characteristics of γc deficiency were linked. Bcl-2 over-expression could correct the abnormal mitochondrial potential in γ CD4+ Ml T cells, and allow these cells to survive in vitro. This improved survival property resulted in a rescue of the peripheral naive T cell lymphopenia observed in the absence of γc23. Nevertheless, the peripheral CD4+ naive T cell numbers were not completely restored (their numbers remained about twofold decreased compared to γc WT mice). Possible explanations for this result include either an incomplete thymic rescue by Bcl-2 12, an additional defect in thymic export and/or the existence of Bcl-2-independent mechanisms for peripheral naive T cell homeostasis.
Considering the different γc-dependent cytokines that are involved in naive T cell homeostasis, a substantial literature exists that provides evidence for a dominant role of IL-7 in naive T cell survival 25, 36, 39. By comparing female Ml mice deficient in γc with those deficient in CD127, we confirm that CD127 ligands are major effectors of naive CD4+ T cell homeostasis. Nevertheless, γc cytokines other than IL-7 are likely to be involved, since a stronger reduction in peripheral T cell numbers was observed in the absence of γc compared to CD127. Which γc-dependent ligands could be involved? We failed to document an effect of CD122 deficiency on naive T cell homeostasis, indicating that IL-2 and IL-15 are redundant in this context. It remains possible that these cytokines become important only in the absence of CD127 although we failed to detect CD122 expression on residual CD4+ T cells in Ml CD127–/– mice (data not shown). Potential roles for IL-4, IL-9 and IL-21 in naive T cell homeostasis therefore remain possible.
Considering Bcl-2-independent mechanisms of γc cytokines, we found that γ Ml CD4+ T cells that have been rescued by Bcl-2 remained small in size. These observations raise questions concerning the potential trophic signals transmitted by γc receptors in naive T cells. Previous work from the Thompson laboratory (reviewed in 30) has enabled the elaboration of a model for cell size regulation. PKB occupies a central role, sitting at the crossroads of the regulation of cell size, metabolism and survival. Activation of PKB results in mTOR-dependent maintenance of glucose and amino acid transporters required for metabolic health of the cell. Nevertheless, the cellular receptor that triggers PKB-dependent trophic effects in naive T cells remains unclear. While IL-7 triggers PKB activation in thymocyte precursors 40 and IL-7 withdraw results in T cell atrophy 40, TCR signals also activate PKB and glucose transporter expression is regulated through the TCR in both thymocyte precursors and peripheral T cells 35, 41. Direct evidence that PKB is required for naive T cell homeostasis is lacking.
We found that constitutive activation of PKB did not increase the size of naive CD4+ T cells in the absence of γc. Therefore, Bcl-2 and PKB activation are not sufficient to maintain naive T cell size and other γc-dependent pathways are involved. In the absence of γc, naive CD4+ T cells expressed low levels of MHC class I molecules and had reduced expression of ribosomal proteins. These observations suggest a potential mechanism through which γc cytokines could exert their effects on cell size. By providing naive T cells with trophic signals that maintain their overall metabolic state (via ribosomal activity), γc cytokines (including IL-7) may assure protein re-synthesis.
While the restoration of Ml CD4+ naive T cell numbers was impressive, it remains to be seen whether other TCR specificities (when placed in the context of γc deficiency) will show the same dependency on Bcl-2. Intrathymically, differentiating CD4 and CD8 SP thymocytes appear to have different requirements for γc cytokines 42. Our preliminary data indicate that Bcl-2 expression can rescue γc deficiency in certain TCR-Tg models that select CD8 T cells (OT-I, P14) but not others (HY-CD8) despite clear Bcl-2 expression in the relevant peripheral T cell population (J.P.D., unpublished observations). Gene expression profiling may provide clues to understand the molecular basis for these differences. It is likely that these observations illustrate the concerted actions of TCR-induced signals and γc cytokine-induced signals in reaching a minimal threshold required for peripheral naive T cell homeostasis 43.
Materials and methods
Ml TCR-Tg mice specific for the male antigen HY Dby on the RAG2–/– or RAG2–/–/γ (C57BL/6) background have been described 23. Mice Tg for an activated PKB 34 or human Bcl-2-Tg mice B6.Cg-Tg(BCL2)25Wehi/J27 were backcrossed to Ml RAG2–/– or RAG2–/–/γ mice to generate ‘monoclonal’ Ml Bcl-2+ mice with or without γc. CD122–/– mice (B6Il2rbtm1Mak/J, JAX) or CD127–/– mice (B6.129S7Il7ratm1Pes/J, JAX) were used to generate ‘monoclonal’ Ml CD122–/– or CD127–/– mice. Mice deficient in RAG2 and IL-7 have been described 20. Donor and recipient mice were used between 4 and 8 wk of age. Animals were kept under pathogen-free conditions in the animal facilities at the Institut Pasteur and all animal experiments were approved by a local committee and in accordance with French law.
Cell isolation and FACS
Single-cell suspensions from thymus and spleen were prepared as described 20. Cell suspensions were stained in PBS with 2% FCS during 15 min on ice in the dark. Before staining, cells were treated with purified mouse IgG to block Fc receptors. Monoclonal antibodies conjugated to fluorescein isothiocyanate, phycoerythrin (PE), PE-Cy5.5, peridinin chlorophyll-a protein-Cy5.5, PE-Cy7, allophycocyanin, allophycocyanin-Cy7, Alexa750 or biotin (eBioscience and BD Biosciences) included CD2 (RM2-5), CD4 (GK1.5, RM4-5), Vβ6 (RR4-7), CD25 (PC61), CD44 (1M7), CD62L (MEL-14), CD69 (H1.2F3), CD122 (TMβ1) and CD127 (A7R34). Biotinylated antibodies were revealed with streptavidin-PE-Cy7 or allophycocyanin-Cy7. Dead cells were excluded using TO-PRO3 or Sytox Green (Molecular Probes). FACS acquisitions were performed using Calibur® or Canto® (BD) analytical flow cytometers, and data sets were analyzed using Flowjo® software.
In vitro T cell activation and proliferation
Splenocytes were labeled with 1 μM CFSE as described 20 and the equivalent of 50 000 CD4+ T cells were co-cultured with an equal number of splenocytes from CD3ϵ–/– male mice in a total volume of 200 μL in RPMI 1640 medium with 10% FCS. Cells were harvested after 4 days and the CFSE dilution profiles of CD4+ T cells assessed by FACS.
Adoptive transfer experiments
Thymocytes from female Ml mice were labeled with CFSE, and the equivalent of 2×106 CD4 SP thymocytes was transferred to unconditioned RAG2–/–/γ or RAG2–/–/γ/IL-7–/– mice (on the C57BL/6 background) via the retro-orbital chamber. Splenocytes were harvested 2 wk later and CFSE+ CD4+ T cells enumerated and analyzed by FACS.
We thank Delphine Guy-Grand, Christian Vosshenrich, Olivia Sappey and Allison Bordack for their helpful discussions and contributions. This work received financial support from the Institut Pasteur, the Institut National de la Santé et de la Recherche Médicale (Inserm), the Ligue Nationale contre le Cancer and from the EC to the MUGEN Network of Excellence LSHG-CT-2005–005203 (to J.P.D.), from the Université Paris 7 and the Fondation de la Recherche Médicale (to G.X.M.) and from FCAAI/Bayer/IRSC (to H.D.). The authors declare having no competing conflicts of interest.