Thymocytes between the β-Selection and Positive Selection Checkpoints Are Nonresponsive to IL-7 as Assessed by STAT-5 Phosphorylation¹

T cell development begins in the bone marrow with the division and self-renewal of hemopoietic stem cells (1, 2). Progression through early maturational events leads to an emigration of a portion of these hemopoietic stem cell descendants from the bone marrow that colonizes the thymus. Ensuing execution of developmental programs, in concert with both intrinsic and extrinsic cues, directs these thymic immigrants along developmental pathways that eventually give rise to the vast majority of T cell subsets (Fig. 1) (3, 4, 5). Central to the generation of these cells is the assembly of a surface-expressed TCR (6). In the case of TCRαβ⁺ thymocytes, there are several checkpoints during thymocyte development in which life/death decisions concerning the developing cell are made based on the expression status of the TCR (or a portion thereof): in-frame TCR β-chain rearrangement (β-selection) (7, 8), in-frame TCR α-chain rearrangement (9), positive selection, and negative selection (10). Transition through these checkpoints is primarily regulated by successful in-frame rearrangement of the TCR β- and α-chains and by the ability of the generated TCR to bind to MHC:peptide with appropriate affinity. Although expression of an appropriate TCR is absolutely required for αβ T cell maturation and function, there are other equally important events that occur during thymocyte development that affect the production of competent T cells, such as the sensitivity of thymocytes toward growth factors and other modulators of cell survival and differentiation (11, 12).

IL-7 is an essential factor that is required for early T cell development. This is apparent because there is a dramatic loss of thymocytes and mature T cells in mice that lack IL-7 or IL-7Rs (13, 14). IL-7 is produced by stromal cells in the bone marrow and by thymic epithelial cells (15, 16, 17, 18). One way that IL-7 exerts its effects on fully mature peripheral T cells, as well as developing T cells in the thymus, is by up-regulating the Bcl-2 family of antiapoptotic proteins and promoting survival (19, 20, 21, 22, 23). IL-7 also influences T cell homeostasis by up-regulating IL-2R expression and thus augmenting cell cycling during activation (24). Although IL-7 was originally thought to be the major thymocyte growth factor (25, 26, 27) and may indeed augment thymopoiesis by promoting proliferation of some subsets, its survival signaling via Bcl-2 induction is also apparently important for thymocyte development. This is evident from experiments with IL-7Rα-deficient mice with forced expression of Bcl-2 that shows rescue of T cell development (28).

IL-7 signaling is initiated by IL-7 binding to a heterodimeric surface receptor complex that consists of a 75-kDa α subunit (29) associated with the common γ-chain (γ_c)³ of the IL-2R family. Following receptor engagement, Jak family kinases (Jak1 and Jak3) associated with the receptor subunits are activated via cross-phosphorylation and in turn phosphorylate STAT-5 docking sites on the receptor (30). STAT-5 is then recruited to the receptor and activated via phosphorylation by the Jak kinases, which enables STAT-5 translocation to the nucleus, where it initiates mRNA transcription. SOCS-3, a member of the suppressors of cytokine signaling family, may also be involved in IL-7/STAT-5 signaling because it has been shown to be a modulator of the IL-2 response through association with or inhibition of Jak-1 and subsequent inhibition of STAT-5 phosphorylation (31, 32). Although not addressed in this work, IL-7R engagement also activates the phosphatidylinositol 3-kinase pathway, resulting in Akt activation (19, 33, 34).

Previous studies have addressed the question of IL-7 signaling during thymocyte development by assessing IL-7Rα expression on defined thymocyte subsets (35, 36). Our approach toward defining IL-7-regulated activity during T cell development has been to evaluate STAT-5 phosphorylation in response to IL-7 in thymocyte subsets by either analysis of intracellular staining by flow cytometry or Western blotting of lysates from sorted populations. Employment of our intracellular staining assay has provided unique advantages by which to gauge the involvement of IL-7 in thymocyte development. For example, we have identified populations that stain poorly for IL-7Rα, yet are responsive as measured by our assay. Additionally, the intensity of the stain has been useful in judging the relative level of the response. In this study, we describe experiments in which we analyzed thymocyte subsets flanking lineage decision and selection checkpoints and found that thymocytes shut off IL-7 responsiveness (as measured by STAT-5 phosphorylation or IL-7-mediated survival) between the β-selection and positive selection checkpoints. Based on the evidence that IL-7 is a key regulator of thymocyte development and our data demonstrating IL-7 nonresponsiveness between major thymocyte selection checkpoints, we hypothesize that regulation of thymocyte responsiveness toward IL-7 is a mechanism by which enforcement of thymocyte death by neglect due to failure of positive selection is mediated.

C57BL/6 female mice were purchased from The Jackson Laboratory (Bar Harbor, ME) or Harlan Sprague-Dawley (Indianapolis, IN). Mice were housed in a U.S. Department of Agriculture-approved facility at the University of Tulsa in accordance with procedures outlined in Guide for the Care and Use of Laboratory Animals (National Research Council) and were all healthy and 4–8 wk of age at the time of thymus harvest. Animal protocols were approved by the Institutional Animal Care and Use Committees of the University of Tulsa (Tulsa, OK) and the University of Oklahoma Health Sciences Center.

Single cell thymocyte suspensions were generated by forceful disruption of thymuses with 3-ml syringe plungers on 70-μm nylon screens. Thymocyte manipulations were conducted in complete tumor medium (37). Thymocytes were further treated with RBC lysis buffer (Sigma-Aldrich, St. Louis, MO), according to manufacturer’s instructions, and then washed into complete tumor medium.

The various fluorochrome- or biotin-conjugated specific Abs and streptavidin conjugates used for thymocyte surface phenotyping were from BD PharMingen (San Diego, CA). Sorting and analysis of thymocyte subsets were performed on MoFlo (DakoCytomation, Fort Collins, CO) and FACSCalibur (BD Biosciences, San Jose, CA) instruments, respectively. MoFlo sort logic was achieved with Summit software (version 3.0), while postsort analysis was performed with CellQuest software. Anti-CD4 FITC and anti-CD8 allophycocyanin were used for discrimination of the four major thymocyte populations: CD4⁻8⁻, CD4⁺8⁺, CD4⁺8⁻, and CD4⁻8⁺ (double negative (DN), double positive (DP), CD4 single positive (SP), and CD8 SP, respectively) in sorting and intracellular staining experiments. To assess the DN subcompartments by flow cytometry for phosphorylated STAT-5 (phospho-STAT-5), thymocytes were labeled with anti-CD4 PerCP, anti-CD8 PerCP, anti-TCRγδ biotin/streptavidin PerCP, anti-CD25 FITC, and anti-CD44 allophycocyanin. DN thymocytes were then obtained by sort purification of the PerCP-negative fraction, followed by evaluation of the DN subcompartments: CD44⁺25⁻, CD44⁺25⁺, CD44⁻25⁺, and CD44⁻25⁻ (DN I-IV, respectively). Sort purification of DN III and IV thymocytes for Western blotting was achieved by exclusion of CD4-, CD8-, TCRγδ-, and CD44-positive thymocytes identified with allophycocyanin-conjugated Abs. Anti-CD25 PE was then used to discriminate DN III (PE-positive) and DN IV (PE-negative) thymocytes. To assess events surrounding positive selection, DN thymocytes were sort depleted from total thymocytes after staining with CD4 PerCP and CD8 PE, and then the remaining thymocytes (DP, CD4 SP, and CD8 SP) were evaluated by staining with TCRβ allophycocyanin- and CD69 FITC-labeled Abs. DP thymocytes for intracellular staining were purified by triple sorting PE/PerCP DP thymocytes labeled with anti-CD4 PerCP and anti-CD8 PE. IL-7Rα expression within the DN compartment was determined by comparison of IL-7Rα PE stains with PE-labeled isotype control stains of the relevant subsets. Gating for these analyses was accomplished by exclusion of CD4-, CD8-, and TCRγδ PerCP-positive cells and evaluation of DN I-IV subsets based on CD25 FITC and CD44 allophycocyanin staining. IL-7Rα expression on post-DN thymocytes was determined by staining thymocytes with PE-labeled IL-7Rα-specific and isotype control Abs with gating based on TCRαβ allophycocyanin and CD69 FITC profiles. In these analyses, DN thymocytes were excluded by gating on thymocytes positive for CD4 and/or CD8, as revealed by staining with PerCP-labeled Abs directed to these molecules.

Before treatment with cytokine, thymocyte populations of interest were incubated at 37°C for 20 min in S-MEM (Invitrogen, Carlsbad, CA) to quiet endogenous signaling. The cells were then incubated in either S-MEM alone or with 25 ng/ml IL-7 (R&D Systems, Minneapolis MN) for an additional 20 min. This dose of IL-7 was used based on previous published titrations with peripheral T cells (37). A dose-response assay was also performed on thymocytes that indicated this was an optimal dose for these cells as well (data not shown). At the end of the incubation, cells were either immediately lysed for analysis of phospho-STAT-5 by Western blot or fixed in preparation for intracellular phospho-STAT-5 staining. For thymocytes slated for analysis by intracellular staining, surface phenotyping was performed at 4°C before performing the cytokine response assay.

Thymocyte populations were surface stained with Abs, treated in the IL-7 response assay, and then prepared for intracellular staining with the Fix & Perm cell permeabilization kit from Caltag Laboratories (Burlingame, CA). Additionally, a 10- to 30-min ice-cold methanol treatment step between fixation and permeabilization of the cells was performed to enhance intracellular staining. It should be noted that it was necessary to sort purify certain populations in some assays before intracellular staining as the methanol treatment step quenches fluorescence of PE and PerCP fluorochromes. Permeabilization of the cells was performed in the presence of either anti-phospho-STAT-5 (Tyr⁶⁹⁴) (Cell Signaling, Beverly, MA) or normal rabbit IgG (Caltag Laboratories) as a control, each at a final concentration of 2.4 μg/ml. Primary Abs were detected with PE-conjugated goat F(ab′)₂ anti-rabbit IgG (Caltag Laboratories) diluted in permeabilization buffer and used at a final concentration of 2 μg/ml. Alternatively, PE-conjugated anti-phospho-STAT-5 (Tyr⁶⁹⁴) (BD PharMingen) or control PE-conjugated mouse IgG1 (BD PharMingen) was used for some experiments. For intracellular P-STAT-5 analysis of direct ex vivo thymocytes, cells were meticulously maintained at 4°C up to the point of fixation, and were not rested or treated with IL-7. All intracellular staining was performed at the indicated concentrations in 100 μl at 5–10 × 10⁷ thymocytes/ml.

Equivalent numbers of sorted thymocytes were treated in the IL-7 response assay, and then 1% Nonidet P-40 lysates were generated, as described previously (37). Lysates were separated in 7.5% SDS-PAGE gels and transferred to polyvinylidene difluoride membranes. After blocking with TBS containing 5% milk and 1% FBS, membranes were incubated either with anti-phospho-STAT-5 (Cell Signaling) Ab, anti-STAT-5 (Santa Cruz Biotechnology, Santa Cruz, CA), or anti-β-actin (Santa Cruz Biotechnology). Typically, total STAT-5 blots were generated from stripped phospho-STAT-5 blots. Blots were further incubated with appropriate peroxide-conjugated secondary reagents, developed with Biowest ECL from Ultraviolet Products (Upland, CA) or ECL from Amersham Biosciences (Piscataway, NJ), and visualized on an Amersham Biosciences VDS-CL Imaging system.

RNA from sorted thymocyte populations was isolated using TRIzol-LS reagent (Invitrogen), according to the manufacturer’s instructions. The RNA was treated with DNase I (Invitrogen) and then reverse transcribed using Invitrogen’s Thermoscript RT-PCR system. Each PCR included 1× PCR buffer with 1 mM MgCl₂ (Invitrogen), 1.25 mM dNTPs (Applied Biosystems, Foster City, CA), 1 U of platinum TaqDNA polymerase (Invitrogen), 0.12% BSA (Sigma-Aldrich), 0.2 μM each custom primer pair (Applied Biosystems), and Sybr Green I (Molecular Probes, Eugene, OR) at a final dilution of 1/20,000. The following primer pairs were used: β-actin forward, 5′-CAGCTTCTTTGCAGCTCCTT-3′; β-actin reverse, 5′-TCACCCACATAGGAGTCCTT-3′; T early α transcript (TEA) forward, 5′-GGACAACCTGGCTTAATGGATACG-3′; TEA reverse, 5′-TTCTCGGTCAACGTGGCATCACAG-3′; CD25 forward, 5′-ACCAGCAACTCCCATGACAAAT-3′; CD25 reverse, 5′-TGAACACTCTGTCCTTCCACGA-3′; CD127 (IL-7Rα) forward, 5′-TGAAAGCAACTGGACGCATGT-3′; CD127 (IL-7Rα) reverse, 5′-ACTTGGCAAGACAGGATCCCA-3′; CD132 (γ_c) forward, 5′-AGATCGAAGCTGGACGGAACT-3′; CD132 (γ_c) reverse, 5′-GCTTCCAGTGCAAACAAGGAA-3′; SOCS-3 forward, 5′-GCGAGAAGATTCCGCTGGTA-3′; SOCS-3 reverse, 5′-CCGTTGACAGTCTTCCGACAA-3′.

PCR was conducted for 40 cycles in a Cepheid Smart Cycler (Sunnyvale, CA). Each cycle included a 15-s denaturing step at 95°C, a 45-s annealing step at 58–66°C (primer dependent), and a 45-s extension step at 72°C. At the conclusion of every run, a melting curve analysis was performed to ensure that only a single product was amplified. Optical data were collected during the annealing step, which were exported into MS Excel, and relative PCR units were calculated and normalized to β-actin. The p values were calculated in GraphPad Prism version 4.

Thymocyte subsets were sort purified and then incubated in 96-well plates (∼1 × 10⁵ thymocytes/well/200 μl) in a humidified incubator at 37°C/5% CO₂ for 24 or 48 h without or with 25 ng/ml IL-7. Thymocytes were evaluated in triplicate at each time point by annexin-V-FITC staining, and the percentages of annexin-V-negative (surviving) cells were expressed as mean ± SD.

The first experiment using our intracellular IL-7 response assay focused on thymocyte subsets defined on the basis of CD4 and CD8. Fig. 2,A depicts surface phenotyping of thymocytes by use of these markers. The relevant gates and corresponding subset identification are indicated. Gated subsets and their STAT-5 phosphorylation status from both −IL-7 and +IL-7 samples are shown in Fig. 2,B. Percentages of cells responding to IL-7 as assessed by STAT-5 phosphorylation are shown within the respective histograms. In the samples treated with IL-7, differential sensitivity was observed between the various subsets. The DN and both SP subsets responded to IL-7 as measured by this assay (69–76%). The DP population responded poorly, however, with ∼1% registering positive for STAT-5 phosphorylation. To validate data obtained by our intracellular phospho-STAT-5 assay, we sorted the major thymocyte populations described above and assessed STAT-5 phosphorylation by Western blotting (Fig. 2 C). Each isolated subset was incubated with either IL-7 (+IL-7) or S-MEM alone as a control (−IL-7). DN thymocytes displayed the highest level of STAT-5 phosphorylation, while the DP thymocytes did not appear to respond. The CD4 SP and CD8 SP thymocytes displayed relatively comparable signals, although as a whole the SP signaling was weaker than that observed in the DN thymocytes. Blotting of β-actin from the same lysates was performed as a loading control. Stripping and reprobing of the STAT portion of the blot with an Ab specific for total STAT-5 indicated the presence of total STAT-5 in all thymic subpopulations. Taken as a whole, the Western blot data corroborated our intracellular results and confirmed differential sensitivity to IL-7 at various stages of thymocyte development.

The observation of the robust phosphorylation of STAT-5 in thymocytes of the DN compartment led to experiments to identify the responsive populations. Sorted DN thymocytes were labeled to discriminate CD25 and CD44 expression (Fig. 3,A) and then subjected to the IL-7 response assay and intracellular staining for STAT-5 phosphorylation. As demonstrated in Fig. 3,B, ∼26% of DN I thymocytes phosphorylated STAT-5 after IL-7 treatment. Passage of thymocytes to the DN II and subsequently the DN III compartments resulted in an increase in IL-7-responsive cells to a point in which virtually all thymocytes were positive (97 and 95%, respectively). In-frame rearrangement and expression of a TCR β-chain in thymocytes at this stage allow for continued development via the process of β-selection. The ensuing propagation of signals initiated by β-selection brings about the cellular changes necessary to prevent death by neglect due to failure of TCRβ rearrangement as well as those required for passage to the next phase of development. As shown in this study, transition from the DN III to the DN IV compartment (β-selection) resulted in a down-regulation of IL-7 sensitivity in DN IV thymocytes, as evidenced by only 39% of this population registering positive for STAT-5 phosphorylation. Comparison of the mean fluorescence intensities from the positive fractions within each of these populations revealed that DN IV thymocytes responded with 50% less intensity than did the DN III thymocytes (data not shown). In sum, the data from this experiment indicate that not only are there fewer responding cells in the DN IV subset, but of the cells that do respond, the response is weaker. Western blot analysis of STAT-5 phosphorylation in lysates from sorted DN III and DN IV thymocytes (Fig. 3 C) indicated that while DN III thymocytes responded quite well to IL-7, the response of the DN IV thymocytes was greatly diminished. This phenomenon cannot be explained by a lack of STAT-5, as the total STAT-5 blots demonstrated the presence of STAT-5 in both populations. Overall, these findings indicate that thymocytes down-regulate IL-7 responsiveness following β-selection.

The loss of IL-7 responsiveness in DN IV thymocytes prompted experiments to evaluate potential mechanisms by which IL-7 sensitivity could be regulated. Comparison of IL-7Rα expression levels by flow cytometry did not indicate a loss of IL-7Rα in the DN IV cells compared with the DN III population (Fig. 4,A). IL-7Rα expression was highest in a subset of the DN I population, reduced in the DN II cells, and reduced to even lower levels in the DN III and DN IV cells. This result was surprising because the DN III cells signaled robustly in response to IL-7 (Fig. 3). The observation of similar IL-7Rα expression in the DN III and DN IV populations is striking when contrasted with the dramatic differential IL-7 responsiveness of these populations and suggests additional mechanisms contribute to the loss of signaling in the DN IV subset.

By relative real-time RT-PCR, we evaluated transcript abundance of the two chains of the IL-7R (IL-7Rα and γ_c) and SOCS-3 in DN III and DN IV thymocytes (Fig. 4 B). Comparison of values from DN III to DN IV thymocytes revealed that the abundance of IL-7Rα transcripts was ∼4.4-fold lower in DN IV in comparison with DN III thymocytes (p = 0.011), while both populations expressed equivalent amounts of the transcript for the γ_c. SOCS-3 is involved in STAT-5 signaling as a negative regulator (31, 32) and potentially as a product induced by IL-7 signaling; therefore, SOCS-3 transcription was assessed in DN III and DN IV thymocytes. This analysis indicated that there was a 17.3-fold decrease (p = 0.016) in SOCS-3 transcript abundance in comparison of DN IV with DN III thymocytes. As controls, we confirmed the composition of our sorted populations by evaluation of CD25 and TEA abundance. As expected, transcription of CD25 decreased (100-fold; p = 0.0001) and TEA transcription increased (from below detectable limits to measurable levels) in comparison of DN III with DN IV thymocytes. SOCS-3 was expressed in the DN III population, indicating that transcription of SOCS-3 correlates with IL-7-responsive rather than nonresponsive populations. In this light, the pattern of SOCS-3 expression observed suggests that 1) SOCS-3 is induced as a result of IL-7 signaling, and 2) SOCS-3 is not responsible for loss of IL-7 responsiveness in β-selected thymocytes.

The loss of IL-7 sensitivity in thymocytes in the immediate post-β-selection period prompted examination of populations surrounding the next thymic checkpoint. DN thymocytes were sort depleted from total thymocyte preparations, and the remaining thymocytes (DP and SP) were evaluated on the basis of TCRβ and CD69. Use of these markers provides a means of identifying populations preparing for (TCR^−/lowCD69⁻), undergoing (TCR^low/intCD69⁺), and emerging from (TCR^high) the positive selection process. Fig. 5,B depicts the phosphorylated STAT-5 levels of the gated populations delimited in Fig. 5,A. Marked absence of IL-7 responsiveness was apparent in the TCR⁻CD69⁻ and TCR^lowCD69⁻ populations, as there was virtually no difference in the phospho-STAT-5 levels between IL-7-treated and control samples. Correlating with the onset of positive selection, a portion of the TCR^lowCD69⁺ thymocyte population exhibited restoration of IL-7 sensitivity, as 4% were positive for STAT-5 phosphorylation. Progression of the TCR^lowCD69⁺ thymocytes to the TCR^int compartment was associated with increased percentages of IL-7-sensitive cells (42%) that reached 78% in the TCR^high populations. The utility of TCR and CD69 expression in evaluating thymocytes undergoing positive selection has been demonstrated in this study and by others (38, 39). Defining thymocytes on the basis of these two markers allows selection events to be followed irrespective of CD4 or CD8 coreceptor expression. However, the demonstration of a small fraction of IL-7-responsive thymocytes within the DP compartment prompted experiments to determine the TCR/CD69 profile of these cells. Triple-sort-purified DP thymocytes (99.9% pure) were subjected to our IL-7 response assay, and nonresponsive and responsive cells were evaluated for TCRβ and CD69 expression. As shown in Fig. 6, the IL-7-nonresponsive DP thymocytes were predominately CD69⁻, while the responsive thymocytes were primarily CD69⁺. This finding indicates that within the DP population, the fraction of thymocytes responding to IL-7 has undergone positive selection.

To further explore the regulation of potential IL-7 responsiveness in thymocytes surrounding the positive selection checkpoint and to contrast the usefulness of our IL-7 response assay with conventional receptor staining, we evaluated IL-7Rα expression on thymocytes transitioning through this phase of development (Fig. 7). DN thymocytes were excluded from the analysis by gating on CD4- and/or CD8-positive thymocytes identified by staining with PerCP-labeled Abs. Similar to our observations with respect to STAT-5 phosphorylation in the IL-7 response assays, the TCR⁻ and TCR^lowCD69⁻ thymocytes were negative for expression of IL-7Rα. A small fraction of the TCR^low/CD69⁺ thymocytes (2%) had restored expression that increased to 19% in TCR^intCD69⁺ thymocytes. Finally, progression to the TCR⁺ stage was associated with detectable expression on 39% of these cells. A comparison of these results with those shown in Fig. 5 demonstrates an advantage of our sensitivity assay, as staining for receptor expression in Fig. 7 underrepresented the proportion of cells capable of responding to IL-7. Taken together, these data suggest that the process of positive selection restores IL-7 signaling that was initially shut off after β-selection.

The observation of differential STAT-5 phosphorylation in response to IL-7 in developmentally defined thymocyte subsets prompted experiments to determine whether populations identified as responsive or nonresponsive would also exhibit differential IL-7-mediated survival signaling, as there are other signaling cascades driven by IL-7. Sorted DN III, DN IV, TCR^−/lowCD69⁻, and TCR^high thymocyte subsets were cultured in vitro in the absence and presence of 25 ng/ml IL-7 and evaluated at 0, 24, and 48 h, and the percentages of surviving cells were calculated as assessed by lack of annexin-V staining. As shown in Fig. 8, enhanced survival of DN III, DN IV, and TCR^high thymocytes was appreciated in cultures containing IL-7. As predicted based on our IL-7 response assays, the survival of DN IV thymocytes was less dramatic than that observed in DN III and TCR^high thymocytes. Indeed, characterization of DN IV thymocytes as a transitional population in which IL-7 sensitivity is waning is further strengthened by the observation that in the next stages of development (TCR^−/lowCD69⁻) thymocytes were observed to be completely insensitive to the protection afforded by IL-7, as evidenced by parallel death rates regardless of IL-7 presence. These results corroborate findings presented above with regard to identification of IL-7-responsive populations and support the proposed role of IL-7 as a regulator of thymocyte death by neglect due to failure of positive selection.

To strengthen our in vitro findings with respect to the contribution of IL-7/STAT-5 signaling in thymocyte development, we evaluated P-STAT-5 levels in freshly isolated thymocytes (Fig. 9). Thymocytes were harvested directly into ice-cold buffer, and cell suspensions were kept cold throughout the assay. Even though multiple cytokines can induce STAT-5 phosphorylation, we found that ex vivo thymocyte subsets show a pattern of STAT-5 phosphorylation similar to that seen in the in vitro IL-7 stimulation assay shown in Fig. 2. The majority of the DN thymocytes stained positive for P-STAT-5, while the P-STAT-5 staining for DP thymocytes was barely detectable. CD4 and CD8 thymocytes were intermediate in their P-STAT-5 levels, with the CD8 thymocytes containing a slightly higher level than the CD4 thymocytes.

By scanning across thymocyte developmental pathways with both intracellular staining and Western blot techniques designed to identify IL-7-responsive populations, we have discovered a phase in thymocyte development in which sensitivity to IL-7 is down-regulated. Interestingly, this stage corresponds with thymocyte differentiation that occurs between the β-selection and positive/negative selection checkpoints and encompasses ∼80% of all thymocytes (Fig. 10). As IL-7 has clearly been shown to be an essential regulator of T cell development (13, 14) coupled with the fact that thymic selection checkpoints exist to elicit properly reactive mature T cells, the demonstrated reciprocal sensitivity of thymocytes toward IL-7 in flanking populations at these checkpoints implies a role for IL-7 signaling in governance of developmental decisions that are primarily based on in-frame Ag chain rearrangements. In-frame rearrangement of TCR β- and α-chains and the ability of generated TCRs to interact with MHC:peptide with appropriate affinity control life/death decisions faced by nascent T cells. Successful rearrangements promote development toward maturity, while failure to do so generates defunct thymocytes that are cleared as a normal part of intrathymic T cell development. Our data suggest that these fates are most likely regulated by the receptiveness, or lack thereof, of thymocytes to IL-7-mediated signaling. Besides demonstrating the advantages of intracellular phospho-STAT-5 staining, our current findings further clarify the role IL-7 plays in regulating T cell development.

We have assessed IL-7 responsiveness within the major thymic populations as traditionally defined by CD4 and CD8 expression (DN, DP, CD4 SP, and CD8 SP). Additionally, we have detected phosphorylated STAT-5 in direct ex vivo thymocyte subsets defined by CD4 and CD8. We have shown that responsiveness can be detected within the DN population as early as the DN I compartment. Differential signaling within this compartment, however, indicates that the DN I thymocytes are a heterogeneous population. The presence of both IL-7-responsive and nonresponsive cells in the DN I compartment is not surprising because Allman et al. (40) recently reported that a subset of this compartment consists of c-Kit^high early T lineage progenitors that lack IL-7Rα. They found these cells to be insensitive to IL-7 survival signaling. They also describe a c-Kit^−/lowIL-7Rα⁺ DN I population that probably represents the most proximal early T lineage progenitor progeny. We too found the DN I compartment to be comprised of both IL-7Rα⁺ and IL-7Rα⁻ cells (Fig. 4 A), which most likely explains the heterogeneity of the IL-7 response.

Thymocyte transition through the next stages of development (DN II and DN III) correlated with the ability of all thymocytes within these populations to phosphorylate STAT-5 in response to IL-7. β-selection (i.e., a positive developmental cue made on the basis of in-frame TCR β-chain rearrangement) begins during this stage of development. Consequences of β-selection include down-regulation of CD25, proliferation, allelic exclusion at the TCR β-chain locus, initiation of TEA transcription, and up-regulation of CD4 and CD8 expression. Our results indicate that down-regulation of IL-7 responsiveness is also a fate experienced by β-selected thymocytes. An unexpected and interesting result was that the dramatic loss of IL-7 responsiveness in the DN IV thymocytes compared with the DN III cells was not due to lower levels of surface expression of IL-7Rα. Loss of IL-7Rα mRNA, however, was already evident in the DN IV population, which is expected because these transitional cells appear to completely lose receptor expression and sensitivity by the next stage of development (DP TCR⁻). The difference in signaling between the DN III and DN IV cells despite similar levels of IL-7Rα suggests additional mechanisms of inhibition of IL-7 signaling are present in the DN IV cells. A recent report by Trigueros et al. (41) described an up-regulation of IL-7Rα on recombination-activating gene-1^−/− thymocytes following administration of anti-CD3ε in vivo. This up-regulation correlated with a transient increase in IL-7Rα transcription, and DN IV thymocytes were found to be dependent on IL-7 for survival. The authors conclude that β-selection promotes survival of the transitioning cells via sustained IL-7 signaling. We found the DN IV transitional population to already be losing IL-7 responsiveness, as measured by the ability to phosphorylate STAT-5. This population, however, still retained a low level of IL-7Rα (Fig. 4,A) and still showed some responsiveness to IL-7 survival signaling (Fig. 8). It is possible that the cells successfully undergoing β-selection transiently maintain IL-7Rα expression, which sustains them for a period of time; however, these cells lose receptor expression and IL-7 responsiveness by the time they reach the DP compartment.

Evaluation of thymocytes that completed β-selection and exited the DN compartment was made on the basis of TCRβ and CD69 expression. The earliest thymocytes in this scheme are TCR⁻, as TCR α-chain rearrangement takes place in this compartment and must be successfully completed before expression of surface TCR. Expression of TCR α-chains in heterodimeric TCRαβ, albeit at low levels, marks successful transition to the TCR^lowCD69⁻ compartment. This population consists of thymocytes that are taxiing in preparation for positive selection should their TCRs prove competent upon encounter of suitable selecting MHC:peptide complexes. Appropriate TCR/MHC:peptide interaction results in positive selection that correlates with CD69 up-regulation (38, 39). Thus, CD69 expression provides a marker by which to identify thymocytes that are auditioning for or have just completed positive selection (38, 39). These TCR^lowCD69⁺ thymocytes further up-regulate TCRαβ expression during progression to SP status. Use of intracellular phospho-STAT-5 staining allowed us to evaluate sequential progenitor-progeny relationships within these populations to contrast the implied IL-7 sensitivity of these populations against current developmental paradigms.

Our results suggest that positive selection initiates restoration of IL-7Rα expression and IL-7 responsiveness and imply that this is the mechanism by which positively selected thymocytes are rescued from death by neglect. The presence of both IL-7-responsive and nonresponsive thymocytes within this population and the low level of TCR expression on these cells, when compared with the profile of the next developmental compartment, suggest additional requirements must be met to restore full IL-7 responsiveness and high-level TCR expression. We have demonstrated by both gating on DP thymocytes in Fig. 2 and purification of DP in Fig. 6 that there is a small population (∼1%) of cells that exhibit IL-7 responsiveness. By use of highly purified DP thymocytes, we have shown that the responsive cells are predominately TCR^low/intCD69⁺ and therefore appear to be cells undergoing/emerging from positive selection. It should be noted that although there is a large decrease in the number of thymocytes in IL-7-deficient mice (compared with wild type), the DP cells in these mice are still capable of differentiating into SP cells (13, 42). This suggests that the function of IL-7 at this stage is not to drive differentiation, but to promote survival. The fact that any thymocytes can survive past the early developmental stages in the absence of IL-7 suggests that other thymic factors are present that are capable of promoting survival, or alternatively that there are thymocytes that are not programmed to die in the same manner as the majority.

The cessation of IL-7 signaling in β-selected thymocytes and subsequent restoration following positive selection suggest an elegant mechanism by which death by neglect at the positive selection checkpoint is regulated. The withdrawal of β-selected thymocytes from the tonic effects of IL-7 may effectively set a clock that marks the beginning of death by neglect that can only be circumvented by restoration of IL-7 signaling. Thymocytes that have passed positive selection once again exhibit sensitivity toward IL-7, thus rescuing them from death by neglect. This model is bolstered by our results demonstrating the gradual decline and then loss of IL-7-mediated survival signaling in thymocytes between β-selection and positive selection. Despite the ability of multiple cytokines to promote STAT-5 signaling, we found that unstimulated ex vivo thymocytes also exhibited the overall pattern of loss of STAT-5 phosphorylation after transition out of the DN stage, followed by recovery upon maturation to the SP stages.

Our findings suggest that thymocyte responsiveness to IL-7 is an additional signal that could participate with TCR-mediated signaling to accomplish appropriate thymocyte maturation programs. The pattern of IL-7 responsiveness that we describe for post-β-selected thymocytes highlights this concept. TCR^lowCD69⁻ thymocytes receive a TCR-induced signal (positive selection) inducing up-regulation of TCR and CD69, followed by a second signal (IL-7 signaling) to effectively carry out the task at hand (rescue from death by neglect). In this model, moderate signaling through the TCR signifies the ability to become a competent T cell; therefore, IL-7 sensitivity is restored to promote survival. This level of regulation achieved by integration of both TCR and IL-7 signals is not unprecedented. Persistence of long-lived peripheral CD4⁺ T cells has been shown to require both MHC interaction as well as IL-7 (17, 43). Therefore, by virtue of dependence on a survival factor present in limiting quantities, homeostatic balance with respect to overall number of T cells can be maintained. The necessity of both types of signals in the maintenance of peripheral homeostasis is most likely an extension of instructional patterning imparted to T cells during development within the thymus. The process of positive selection could then be viewed as a required step necessary before selecting thymocytes based on their ability to respond to IL-7. In this light, positive selection then generates a pool of thymocytes that compete for IL-7, which in turn gives rise to those cells that are most capable of benefiting from this cytokine. Taken as a whole, the data presented in this work demonstrate a role for IL-7 in enforcement of decisions reached by developing thymocytes at both the β-selection/lineage decision and positive/negative selection checkpoints.

We are grateful to Dr. Thomas A. Broughan for providing departmental funding in support of this project. We also thank Debbie Neal for compliance and secretarial support. We thank Dr. Linda F. Thompson at the Oklahoma Medical Research Foundation for critical review of the manuscript.