Deficiency of the Src Homology Region 2 Domain ... - Semantic Scholar

The Journal of Immunology

Deficiency of the Src Homology Region 2 Domain-Containing Phosphatase 1 (SHP-1) Causes Enrichment of CD4ⴙCD25ⴙ Regulatory T Cells1 Jennifer D. Carter,* Gina M. Calabrese,* Makoto Naganuma,† and Ulrike Lorenz2* A subpopulation of T cells, named regulatory T cells (Treg cells), has been shown to play a key role in tolerance and the prevention of autoimmunity. It is not known how changes in TCR signal strength during thymic T cell development affect the generation of a Treg population. In this study, we took two different strategies to modulate the TCR signal strength: an intrinsic approach, where signaling was enhanced by the loss of a negative regulator, and an extrinsic approach, where signaling strength was altered through variations in the concentrations of the selecting peptide. The tyrosine phosphatase Src homology region 2 domaincontaining phosphatase 1 (SHP-1) is a known negative regulator of TCR-mediated signaling. motheaten mice, lacking expression of SHP-1, showed a 2- to 3-fold increase in the percentage of CD4ⴙCD25ⴙ Treg cells within the CD4ⴙ T cells. Similarly, the percentage of Treg cells was heightened in fetal thymic organ cultures (FTOCs) derived from motheaten mice compared with wild-type FTOCs, thus establishing the thymic origin of these Treg cells. Using FTOCs derived from DO11.10 TCR transgenic mice, we demonstrated that exposure to increasing concentrations of the cognate OVA peptide favored the appearance of Treg cells. Our data suggest that the development of CD4ⴙCD25ⴙ Treg cells is intrinsically different from non-Treg cells and that Treg cells are selectively enriched under conditions of enhanced negative selection. Our data also reveal a key role for the SHP-1mediated regulation of TCR signal strength in influencing the ratio of Treg vs non-Treg cells. The Journal of Immunology, 2005, 174: 6627– 6638.

I

mmune tolerance and the prevention of autoimmune diseases are thought to occur at various levels in the immune system. During thymic T cell development, the majority of thymocytes expressing TCRs with high affinity for self-peptides undergo negative selection, resulting in their apoptosis. This process is referred to as central tolerance (reviewed in Ref. 1). Despite this mode of eliminating autoreactive T cells, every healthy adult still carries potentially harmful self-reactive T cells (2). In the periphery, such autoreactive T cells are rendered inert by mechanisms termed peripheral tolerance (3). In addition, a subpopulation of T cells, named regulatory T cells (Treg cells),3 has been thought to play a key role in tolerance and the prevention of autoimmunity (reviewed in Ref. 4). Although the presence of such Treg cells had been proposed for many years (5, 6), their existence was controversial until recently. One type of Treg cell, defined by its surface expression of CD4 and the IL-2R ␣-chain CD25 (IL-2R␣), is now well-accepted as a functionally suppressive T cell subpopulation

*Department of Microbiology and The Beirne Carter Center for Immunology Research, and †Department of Internal Medicine, University of Virginia, Charlottesville, VA 22908 Received for publication November 30, 2004. Accepted for publication March 14, 2005. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by National Institutes of Health Grant RO1 AI48672 (to U.L.). M.N. was supported by the Crohn’s and Colitis Foundation of America and Grants RR00175-01 and DK 56703 to Peter Ernst. 2 Address correspondence and reprint requests to Dr. Ulrike Lorenz, Department of Microbiology, Jordan Hall 7212, University of Virginia Health System, P.O. Box 800734, Charlottesville, VA 22908-0734. E-mail address: [email protected] 3

Abbreviations used in this paper: Treg cell, regulatory T cell; Teff cell, effector T cell; Tg, transgenic; SHP-1, Src homology region 2 domain-containing phosphatase 1; FTOC, fetal thymic organ culture; SP, single positive; DP, double positive; DN, double negative.

Copyright © 2005 by The American Association of Immunologists, Inc.

(reviewed in Ref. 7). Treg cells have been shown to suppress the proliferation and function of effector T cells (Teff cells) both in vitro and in vivo (4). In several mouse model systems, it has been demonstrated that functional Treg cells are involved in the suppression of an autoimmune response, whereas a depletion of this population promotes the development of autoimmune diseases (8 –10). The precise developmental pathway of Treg cells is still being defined. Treg cells are thought to arise primarily in the thymus during the selection process based on the following observations. First, Treg cells are present in the postnatal thymus at day 3 when they cannot be detected in the periphery. Moreover, thymectomy at day 3 results in autoimmune disease (5), which can be prevented by the transfer of CD4⫹CD25⫹ Treg cells (8). Second, in TCR transgenic (Tg) mice, exposure of the developing T cells to the cognate peptide in the thymus caused an increase in the CD4⫹CD25⫹ Treg cell population (11–13). Interestingly, exposure to low-affinity peptides failed to induce an increased Treg cell population. Third, expression of the specific peptide in the cortical epithelium resulted in an increase of CD25⫹ Treg cells (14, 15), suggesting that radioresistant cells can direct CD25⫹ Treg cell development (11). Fourth, in contrast, expression of the high-affinity cognate peptide in bone marrow-derived cells promoted mostly the development of CD25⫺ T cells with regulatory function (14). Based on these and similar studies, it has been hypothesized that Treg cells are selected in the thymus by high-avidity interactions with a cognate peptide. However, because the expression of highavidity peptides in a transgenic TCR system can also promote negative selection, the intracellular TCR signals leading to the distinction between deletion and selection of CD4⫹CD25⫹ Treg cells remain poorly understood. The functional definition of the CD4⫹CD25⫹ Treg cell is its ability to suppress proliferation and effector functions of other T cells while itself being anergic upon physiological, suboptimal stimulation. Interestingly, strong activation of responder cells, e.g., 0022-1767/05/$02.00

6628 via cross-linking of the TCR/CD3 complex or costimulation of CD28, renders them refractory to suppression (16). Although anergy is a hallmark of Treg cells in vitro, they are capable of considerable expansion in vivo without losing their suppressive ability (12, 17). Treg cells must be stimulated to exert their suppressive effects; however, once stimulated, Treg cells act in an Ag-nonspecific manner (8, 9, 18, 19). In vitro, the suppressive abilities of the Treg cells are connected to the anergic state, because addition of IL-2 not only breaks the anergy of the Treg cells but also abrogates their ability to suppress proliferation (20). Src homology region 2 domain-containing phosphatase 1 (SHP-1) is a non-transmembrane protein tyrosine phosphatase that is expressed predominantly in hemopoietic cells of all lineages and all stages of maturation as well as at lower levels in epithelial cells (21–25). The existence of a murine genetic model for SHP-1 deficiency has significantly aided our understanding of the biological function of SHP-1 (26, 27). A splicing mutation in the SHP-1 locus causes the motheaten (me/me) phenotype. This mutation leads to a frameshift near the 5⬘-end of the SHP-1 coding sequence, resulting in no detectable SHP-1 protein. me/me mice are therefore effectively SHP-1 nulls. A number of studies have suggested that SHP-1 is a negative regulator of TCR-mediated signaling (28, 29). We had previously reported that SHP-1 plays a role during thymic T cell development by influencing the TCR signal strength (30). Using a transgenic TCR system, bred into the motheaten background, we observed that SHP-1 is involved in setting the thresholds for positive and negative selection (31). Other reports have also noted similar effects of SHP-1 on thymic selection (32–34). Taken together, these studies are consistent with a regulatory role for SHP-1 during T cell development. In this study, we determined whether SHP-1 plays a role in the development of Treg cells. me/me mice show an increase in the percentage of functional CD4⫹CD25⫹ Treg cells compared with their ⫹/⫹ littermates in the thymus and spleen. These SHP-1deficient Treg cells are thymus derived and functionally indistinguishable from wild-type Treg cells based on their suppressive efficiency and expression of a number of surface markers. Moreover, in fetal thymic organ cultures (FTOCs) derived from TCR-Tg mice, addition of the cognate at concentrations that induce negative selection results in an enrichment of CD4⫹CD25⫹ T cells within the CD4⫹ single-positive (SP) population. These data suggest that the development of CD4⫹CD25⫹ Treg cells is intrinsically different from Teff cells. Our data also provide evidence that the tyrosine phosphatase SHP-1, through its regulation of signaling strength downstream of the TCR and its influence on negative selection, favors the development of Teff cells over Treg cells.

Materials and Methods Mice (BALB/c:me/⫹), (BALB/c:⫹/⫹:DO11.10 TCR-Tg⫹), and (BALB/c:me/ ⫹:DO11.10 TCR-Tg⫹) mice (31) were bred in our colony to generate the various genotypes. Genotyping for all mice was performed as described (31). Unless otherwise noted, 15- to 19-day-old mice were used throughout the study. All mice were bred and maintained in accordance with the policies of the Institutional Animal Care and Use Committee (IACUC) at the University of Virginia. All experiments involving mice were conducted with the approval of IACUC.

Electronic cell purification Splenocytes were dispersed from the spleen, and RBC were lysed using PharM Lyse (BD Pharmingen). Cell suspensions were precleared of macrophages and B cells by adherence to tissue culture plates (Corning) at 1 ⫻ 108 cells/plate for 1 h at 37°C followed by incubation onto a nylon wool column (Polysciences) (4 ⫻ 108 cells/g wool) for 1 h at 37°C. Cells were subsequently stained with anti-CD4-PE and anti-CD25-allophycocyanin, (used at 2 ␮g/ml each) (BD Pharmingen). CD4⫹CD25⫹ and CD4⫹CD25⫺

DEFICIENCY OF SHP-1 CAUSES ENRICHMENT OF Treg CELLS populations were electronically purified using a FACSVantage SE cell sorter (BD Biosciences).

Flow cytometry Total splenocytes or thymocytes were isolated and stained with Abs to the indicated surface markers in PBS supplemented with 1% BSA and 0.1% sodium azide. CD4-PE, CD25-allophycocyanin, CD25-PE, CD25-biotin, CD44-allophycocyanin, streptavidin-PerCP, CD69-PE (used at 1 ␮g/ml), CD8-FITC, CD8-allophycocyanin, V␤8-FITC, V␤8-biotin (clone MR5-2), V␣2-FITC, CD45RB-PE, CD62L-PE, CD38-PE, TLR4-PE, CTLA-4-PE (used at 2.5 ␮g/ml), and streptavidin-allophycocyanin (used at 0.25 1 ␮g/ ml) were purchased from BD Pharmingen, KJ1-26-PE and KJ1-26-FITC (used at 1 ␮g/ml) were purchased from Caltag, and GITR-biotin (used at 0.075 ␮g/ml) was purchased from R&D Systems. Stained cells were collected on a FACSCalibur instrument calibrated with CaliBRITE beads using CellQuest software for collection and subsequent analyses (BD Biosciences). Analyses were conducted on live cells (⬎95%) as defined by forward- and side-angle scatter. Gates were set using isotype-matched control Abs.

Proliferation and suppression assays These assays were performed as described previously (20, 35). Briefly, to assess proliferation, 2.5 ⫻ 104 CD4⫹CD25⫺ or CD4⫹CD25⫹ T cells (electronically sorted to a purity of ⬎95% as described above) were plated in triplicate in 200 ␮l of RPMI 1640 medium (supplemented with 10% FCS, 5 ⫻ 10⫺5 M 2-ME, 2 mM L-glutamine, and antibiotics) in 96-well plates. For assessment of suppressor activity, 2.5 ⫻ 104 CD4⫹CD25⫺ (effector) T cells plus 2.5 ⫻ 104 CD4⫹CD25⫹ (regulatory) T cells were plated in triplicate in 96-well plates. Irradiated (2000 rad) total splenocytes (from a DO11.10 TCR-Tg BALB/c mouse) were added at 5 ⫻ 104 cells per well along with anti-CD3 Ab (145-2c11; Southern Biotechnology) at 12 ␮g/ml. Cells were incubated at 37°C for 72 h before they were pulsed with 1 ␮Ci of [3H]thymidine for 18 h. [3H]Thymidine incorporation was measured using a cell harvester and Betaplate counter (Wallac).

Expression of Foxp3 Total RNA was isolated from electronically purified CD4⫹CD25⫹ or CD4⫹CD25⫺ T cells using TRIzol reagent (Invitrogen Life Technologies). Purified RNA (from 4 ⫻ 104 cells/sample) served as template for firststrand cDNA synthesis using the SuperScript System (Invitrogen Life Technologies). Foxp3 levels were assessed using real-time PCR as described previously (36). For conventional PCR, products generated from 30-cycle reactions were analyzed. The primers for Foxp3 were the same as used for real-time PCR. The following primers were used for GAPDH assessment: 5⬘-GGC GTC TTC ACC ACC ATG GAG-3⬘ and 5⬘-AAG TTG TCA TGG ATT GAC CTT GG-3⬘.

Immunoprecipitation and immunoblotting These assays were performed as described previously (31). Briefly, SHP-1 was immunoprecipitated from lysates of the indicated numbers of purified CD4⫹CD25⫹ or CD4⫹CD25⫺ T cells using 2 ␮g of rabbit anti-SHP-1 Abs (Santa Cruz Biotechnology) followed by immunoblotting for SHP-1 with monoclonal anti-SHP-1 (clone 1SH01; NeoMarkers) at 1.35 ␮g/ml.

FTOCs FTOCs were set up as previously described (37, 38). Briefly, females were checked for fertilization by the presence of a vaginal plug (day 0 gestation) and sacrificed on day 15 gestation when pups were harvested to remove fetal thymi. Thymi were cultured in six-well plates on a Transwell insert resting upon 1.5 ml of Iscove’s medium supplemented with 10% FCS (HyClone), 5 ⫻ 10⫺5 M 2-ME, 2 mM L-glutamine, and antibiotics. Days of subsequent culture are indicated for each experiment. Thymocytes were dispersed, counted, and stained for flow cytometric analysis as described above.

Peptide pulsing of FTOCs FTOCs were established from day 15 gestation pregnant mice as described above. On day 2 of culture, FTOCs were incubated with OVA peptide (ISQAVHAAHAEINEAGR; synthesized at the Biomolecular Research Facility, University of Virginia) at concentrations indicated in the individual experiment. After 5 days of culture, organ cultures were transferred to OVA-free medium and cultured for an additional 48 h without OVA. Thymocytes were then dispersed, counted, and stained for flow cytometric analysis as described above. We observed that ⫹/⫹ FTOCs derived from crosses of me/⫹:DO11.10 mice (as shown in Fig. 7) failed to show a

The Journal of Immunology significant increase in the percentage of CD4⫹CD25⫹ cells until exposed to 2000 ng of OVA peptide, whereas the FTOCs derived from ⫹/⫹: DO11.10 mice (as shown in Fig. 6) start to respond at ⱖ1000 ng of peptide.

Results Thymi and spleens of me/me mice are enriched for CD4⫹CD25⫹ T cells Because high-avidity peptide:TCR interactions can lead to an increase in the number of Treg cells (11–13), it is thought that the strength of TCR signaling also regulates Treg cell development. We therefore examined whether SHP-1, through regulation of TCR signal strength, is involved in the generation of Treg cells. In the descriptions below, we refer to and discuss the absolute and relative cell numbers separately, because the motheaten mice are smaller in size, which in turn affects their thymic cellularity. Moreover, differences in relative and absolute cell numbers have also been useful in distinguishing between possible models of Treg cell generation. Thymi from SHP-1-deficient me/me mice showed consistently higher percentages of CD4⫹CD25⫹ T cells compared with their ⫹/⫹ littermates (4.2 vs 1.8%; p ⬍ 0.0013) (Fig. 1a and

6629 Table I). The absolute number of CD4⫹CD25⫹ T cells in the me/me mice is approximately the same as in their ⫹/⫹ littermates, despite the smaller size of the me/me mice and decreased total thymic cellularity. Importantly, there are only very few (⬍1%) CD25⫹ cells within the double-positive (DP) and the CD8 SP subpopulations in both me/me and ⫹/⫹ thymi, indicating that there is no overall increase in CD25⫹ surface expression in the me/me mice (Fig. 1a). Thymocytes that have undergone positive selection up-regulate the activation marker CD69 (39). In both the me/me and the ⫹/⫹ population, the majority of CD4⫹CD25⫹ and CD4⫹CD25⫺ are CD69high, indicating that they recently underwent positive selection (Fig. 1a), whereas the CD69neg/low population represents the more mature CD4 SP thymocytes. Spleens of me/me mice also have ⬃2-fold more CD4⫹CD25⫹ T cells than their ⫹/⫹ littermates, both in absolute number and as a percentage of the CD4⫹ cells (20 vs 10%, p ⬍ 0.001) (Fig. 1b and Table I). Because CD25 is also a marker for activated T cells, we examined whether there was a general activation of T cells in the me/me mice; however, ⬍2% of the splenic CD8⫹ T cells are

FIGURE 1. me/me mice have higher percentages of CD4⫹CD25⫹ Treg cells than ⫹/⫹ mice. a and b, Thymocytes (a) and splenocytes (b) from ⫹/⫹ and me/me BALB/c mice were isolated and stained for flow cytometry with the indicated Abs. A total of 106 live cells was analyzed per point. Gated subpopulations (based on CD4 and CD8 expression) were further analyzed for CD25 expression and the percentages of CD25⫹ cells within each subpopulation are shown. Overlay histograms of CD69 surface expression of ⫹/⫹ (open line) and me/me (gray shaded) on CD4⫹CD25⫹ as well as CD4⫹CD25⫺ T cells are depicted below. c, CD4⫹CD25⫹ splenocytes isolated from ⫹/⫹ and me/me mice are anergic and suppress the proliferation of Teff cells. CD4⫹CD25⫹ (Treg) and CD4⫹CD25⫺ (Teff) cells were isolated from splenocytes. A total of 2.5 ⫻ 104 cells of each indicated cell type was assayed for proliferation/suppressor activity. Proliferation of ⫹/⫹ CD4⫹CD25⫺ is set as 100%. Where indicated, IL-2 (30 u/ml) was added. APCs represent background proliferation from the irradiated splenocytes. Error bars denote ⫾ SEM. Results are representative of five independent experiments.

6630

DEFICIENCY OF SHP-1 CAUSES ENRICHMENT OF Treg CELLS

Table I. Relative and absolute cell numbers in wild-type and SHP-1-deficient micea

n

% of CD4⫹CD25⫹ T Cells within CD4⫹ Population

No. CD4⫹ T Cells (⫻106)

No. CD4⫹CD25⫹ T Cells (⫻106)

Thymus ⫹/⫹ me/me DO11:⫹/⫹ DO11:me/me

13 9 20 20

1.8 ⫾ 1.4 4.2 ⫾ 1.6b 2.8 ⫾ 1.0 7.1 ⫾ 2.3c

9.9 ⫾ 3.9 4.4 ⫾ 4.0 11.6 ⫾ 4.4 8.7 ⫾ 2.3

1.8 ⫾ 1.3 1.9 ⫾ 0.5e 3.2 ⫾ 1.0 6.2 ⫾ 1.9d

Spleen ⫹/⫹ me/me DO11:⫹/⫹ DO11:me/me

18 16 27 20

10.4 ⫾ 2.3 19.9 ⫾ 5.5c 6.5 ⫾ 2.7 20.2 ⫾ 6.2c

3.2 ⫾ 0.8 3.2 ⫾ 1.3 5.1 ⫾ 1.6 4.1 ⫾ 1.3

3.3 ⫾ 0.8 6.4 ⫾ 1.7d 3.3 ⫾ 1.4 8.3 ⫾ 2.5d

a Thymi and spleens of mice carrying the indicated genotypes were analyzed for CD4, CD8, and CD25 surface expression. The relative (%) and absolute (no.) of cells displaying the indicated phenotypes are shown. Errors represent the SEM. n indicates the number of mice analyzed to determine percentage of CD4⫹CD25⫹ within the CD4⫹ population. b– d p values were calculated for the changes in the absolute numbers of CD4⫹CD25⫹ between me/me and ⫹/⫹ mice. bp value of 0.0013, cp value of ⬍0.001, and dp value of ⬍0.005 at a 95% confidence interval. e Although there is a significant increase in the relative number of CD4⫹CD25⫹ thymocytes in the me/me mouse, the absolute numbers are comparable. However, the cell numbers are not normalized for the size difference between a me/me and a ⫹/⫹ thymus, although a me/me thymus is approximately one-half to two-thirds the size of a ⫹/⫹ thymus (as also reflected in the number of CD4 SP thymocytes). Based on thymic composition and surface marker profile of the me/me thymus (Ref. 28 and Fig. 1), the decreased thymic cellularity is not due to increased deletion but rather reflects the smaller body size of the animals. The actual increase in thymic CD4⫹CD25⫹ T cells in me/me mice is therefore likely underestimated.

CD25⫹. In addition, very few splenic CD4⫹CD25⫹ T cells express the activation marker CD69 both in me/me and ⫹/⫹ mice (Fig. 1b), indicating that the increased fraction of CD4⫹CD25⫹ T cells in the motheaten mice is not due to a general increase in activated T cells. To examine the functional properties of splenic CD4⫹CD25⫹ T cells, in vitro proliferation and suppression assays were performed (Fig. 1c). Electronically sorted CD4⫹CD25⫺ cells (Teff cells) from both ⫹/⫹ and me/me mice proliferate when cultured in the presence of soluble CD3 Ab and APCs. However, CD4⫹CD25⫹ T cells from mice of either genotype did not proliferate under these conditions unless exogenous IL-2 was added, indicating that they are anergic as described for CD4⫹CD25⫹ Treg cells (9). When mixed in culture, CD4⫹CD25⫹ T cells from both ⫹/⫹ and me/me mice were able to suppress the proliferation of CD4⫹CD25⫺ cells from ⫹/⫹ mice, further indicating that the splenic CD4⫹CD25⫹ cells isolated from both ⫹/⫹ and me/me mice contain functional Treg cells (see additional characterization below). These data suggest that me/me mice are enriched for CD4⫹CD25⫹ Treg cells. DO11.10 TCR-Tg me/me mice are enriched for CD4⫹CD25⫹ T cells in their thymus and spleen The role of SHP-1 in regulating thymic T cell selection is best revealed when the me/me genotype is crossed onto a TCR-Tg background (31–34). Therefore, we examined mice expressing the DO11.10 Tg TCR in a me/me background (31) for the presence of CD4⫹CD25⫹ Treg cells. DO11.10 mice express a Tg TCR derived from the DO11.10 T cell hybridoma line that is specific for a fragment of chicken OVA (OVA; aa 323–339) in the context of I-Ad (30). Thymi from DO11:me/me mice contain a greater percentage and absolute number of CD4⫹CD25⫹ T cells than DO11: ⫹/⫹ mice (7.1 vs 2.8%; p ⬍ 0.001) (Fig. 2a and Table I). There is not a general increase in CD25 surface expression in other thymic subpopulations, such as the CD4⫹CD8⫹ DP cells (Fig. 2a). In the spleen, DO11.10 TCR-Tg mice have previously been reported to have fewer CD4⫹CD25⫹ Treg cells (⬃5%), compared with nonTCR-Tg mice (up to 10%) (40). DO11:me/me mice contain more splenic CD4⫹CD25⫹ (13–20% of CD4⫹ population) cells compared with DO11:⫹/⫹ mice (⬃6%) (Fig. 2b and Table I). Very few splenic CD8⫹ T cells in both ⫹/⫹ and me/me mice are CD25⫹ and very few splenic CD4⫹CD25⫹ T cells express the

activation marker CD69 (Fig. 2b), arguing against a general T cell activation. Naturally occurring CD4⫹CD25⫹ Treg cells from DO11.10 TCR-Tg mice have previously been reported to require the expression of an endogenous TCR␣ chain. Although the necessity for the second V␣ chain is not fully understood and has even been shown to be superfluous under certain conditions (12, 13), several nontransgenic V␣ chains have been reported to be enriched in the Treg cell population (40). V␣2 has been found to be expressed on 11– 15% of the DO11:⫹/⫹ Treg cells. The same extent (10 –12%) of thymic CD4⫹CD25⫹ cells in both DO11:me/me and DO11:⫹/⫹ mice express the V␣2 chain, supporting the correlation with Treg cells (Fig. 2c). This enrichment of V␣2 is confined to the CD4⫹CD25⫹ population, because the CD25⫺ populations of CD4⫹ cells in the me/me and control background have a relatively low usage of the V␣2 chain (⬃4%) (Fig. 2c). Although there is an enrichment of V␣2 usage in the splenic CD4⫹CD25⫹ cell population, there is also an increase in V␣2 usage (up to 9%) in the CD25⫺ population due to a general increase in the percentage of T cells expressing an endogenous TCR instead of the transgenic ␣␤TCR (Fig. 2d). Both thymic and splenic CD4⫹CD25⫹ T cell populations still express the transgenic TCR on their surface, detected using the clonotypic Ab KJ1-26 (data not shown). The similarities of surface marker expression profiles between the ⫹/⫹ and me/me CD4⫹CD25⫹ populations and the consistencies with previous reports for CD4⫹CD25⫹ Treg cells further indicate that the CD4⫹CD25⫹ T cells found in the me/me mice are Treg cells. Although there are no unique Treg specific surface markers available at present, we analyzed the CD4⫹CD25⫹ and the CD4⫹CD25⫺ populations from both ⫹/⫹ and me/me mice for the expression of a number of surface markers that have been associated with the Treg population, such as GITRpos, CD45RBlow, CTLA-4pos, CD62Llow, CD38high, and TLR4pos (41– 45) (data not shown). The majority of cells within the CD4⫹CD25⫹ subpopulation in me/me and control mice displayed comparable flow cytometric profiles similar to what has been reported for Treg cells. Moreover, the thymic and splenic CD4⫹CD25⫹ populations from both ⫹/⫹ and me/me mice expressed slightly lower CD4 levels on their surface as it has been described for natural Treg cells (46) (Fig. 2e). These data further indicated that the me/me and the ⫹/⫹


6631

FIGURE 2. me/me DO11.10 TCR-Tg mice have higher percentages of CD4⫹CD25⫹ Treg cells than ⫹/⫹ DO11.10 TCR-Tg mice. a and b, Thymocytes (a) and splenocytes (b) from ⫹/⫹ and me/me DO11.10 TCR-Tg mice were isolated and stained with the indicated Abs. Approximately 106 live cells were collected per point. Gated subpopulations (based on CD4 and CD8 expression) were further analyzed for CD25 expression, and the percentages of CD25⫹ cells within each subpopulation are shown. Overlay histograms of CD69 surface expression on splenic CD4⫹CD25⫹ as well as CD4⫹CD25⫺ T cells from ⫹/⫹ (open line) and me/me (gray shaded) are depicted. c and d, Profiles of V␣2 surface expression of ⫹/⫹ and me/me CD4⫹CD25⫹ and CD4⫹CD25⫺ SP thymocytes (c) and splenocytes (d) are shown. Percentages of V␣2-positive cells within each subpopulation are indicated. e, Density plots of CD4⫹ SP thymocytes and splenic CD4⫹ T cells depicting CD4 surface expression levels. f, CD4⫹CD25⫹ splenocytes isolated from ⫹/⫹ and me/me DO11.10 TCR-Tg mice are anergic and suppress the proliferation of Teff cells. CD4⫹CD25⫹ (Treg) and CD4⫹CD25⫺ (Teff) cells were isolated from splenocytes. A total of 2.5 ⫻ 104 cells of each indicated cell type was assayed for proliferation/suppressor activity. Proliferation of ⫹/⫹ CD4⫹CD25⫺ T cells is set as 100%. Error bars denote ⫾SEM. Results are representative of four independent experiments.

CD4⫹CD25⫹ T cells are comparable, albeit selectively increased in the motheaten mice. Recently, expression of Foxp3, a transcription factor, has been shown to be restricted to Treg cells. Moreover, expression of Foxp3 in T cells seems to be necessary and sufficient to induce the regulatory phenotype (47– 49), thereby making it a valuable marker for naturally occurring Treg cells. Sorted splenic CD4⫹CD25⫹ T cells from both wild-type and me/me mice showed comparable Foxp3 mRNA expression as assessed by quantitative real-time PCR (Fig. 3a) as well as conventional PCR (b). Under the same conditions, the CD4⫹CD25⫺ T cell populations contained only background or below-detection levels of Foxp3 mRNA. These data confirmed that the CD4⫹CD25⫹ T cell populations from wild-type and me/me mice are comparable and also further supported that they are Treg cells. Functionality of sorted splenic CD4⫹CD25⫹ T cells was assessed in suppression assays. CD4⫹CD25⫹ T cells (Treg) from both DO11:⫹/⫹ and DO11:me/me mice failed to proliferate in response to stimulation with soluble anti-CD3 Ab in the presence of APCs. They also suppressed the proliferation of purified CD4⫹CD25⫺ cells (Teff cells) upon coculturing, further indicating that these CD4⫹CD25⫹ cells are functional Treg cells (Fig. 2f).

Splenic CD4⫹CD25⫹ Treg cells from wild-type and me/me mice have comparable suppressive capabilities Although the data presented above suggest that the lack of SHP-1 results in an increase in the percentage of functional CD4⫹CD25⫹ Treg cells, the assays did not address whether there were any differences in the suppressive capabilities of ⫹/⫹ and me/me Treg cells. Assessment of the expression level of SHP-1 in electronically sorted splenic CD4⫹CD25⫺ and CD4⫹CD25⫹ cells showed that SHP-1 is expressed at comparable levels in CD4⫹CD25⫹ Treg cells and CD4⫹CD25⫺ T cells (Fig. 4a). To examine the respective suppressive potentials, we used decreasing ratios of Treg cells to responder cells (Teff cells) in an in vitro suppression assay. Treg cells from both DO11:⫹/⫹ and DO11:me/me mice completely suppressed Teff cell proliferation at a ratio of 1:2, partially suppressed at a ratio of 1:4, and failed to show suppression at dilution of 1/8 (Fig. 4b). At the highest dilution of Treg cells, we noticed that the proliferation was higher than that of Teff cells alone. This may be due to IL-2 produced by the activated Teff cells, which helps to overcome the anergy of the Treg cells in coculture allowing them to also proliferate. Thus, although there are a greater number of Treg cells in the me/me background, the suppressive

6632


FIGURE 3. CD4⫹CD25⫹ Treg cells from wild type and me/me express Foxp3 mRNA. cDNAs were prepared from electronically purified splenic CD4⫹CD25⫹ and CD4⫹CD25⫺ cells (from wild-type and me/me DO11.10 TCR-Tg mice) and assessed for Foxp3. a, Foxp3 mRNA levels were quantitated using real-time PCR. Foxp3 levels were normalized to 18S RNA expression and relative Foxp3 values are shown. b, Using conventional PCR, Foxp3 was amplified at detectable levels from CD4⫹CD25⫹ T cells but not from CD4⫹CD25⫺ T cells. GAPDH is amplified as input control for the cDNA.

potential of individual cells appears comparable between ⫹/⫹ and me/me Treg cells based on the above assays. CD4⫹CD25⫹ T cells are thymus derived It has been reported previously that Treg cells are thymus derived and can be earliest detected in the thymus at day 3– 4 postnatal, whereas they are still undetectable in the spleen at that time (11, 50). Both ⫹/⫹ and me/me mice have CD4⫹CD25⫹ T cells in their thymi by day 4, but me/me mice have a ⬃2-fold higher percentage of CD4⫹CD25⫹ T cells within their CD4⫹ SP population (Fig. 5a). However, neither the ⫹/⫹ nor the me/me mice have detectable CD4⫹CD25⫹ T cells in their spleens. Essentially identical results were obtained when DO11.10 TCR-Tg mice in the ⫹/⫹ and me/me background were analyzed (Fig. 5b). To further test the thymic origin of the CD4⫹CD25⫹ T cells, we used FTOCs. CD4⫹CD25⫹ T cells are readily detectable within the CD4⫹ SP population in day 7 FTOCs derived from ⫹/⫹ and me/me DO11.10 mice (Fig. 5c). FTOCs from me/me mice contain 2- to 3-fold higher percentages of CD4⫹CD25⫹ T cells within their CD4 T cells populations, consistent with the findings in the neonatal and 2- to 3-wk-old mice presented above. Thus, the CD4⫹CD25⫹ T cells detected in the ⫹/⫹ and me/me mice are of thymic origin, and the increased percentages detected in me/me mice are already present very early in the life of the mouse. This further supports the hypothesis that the increase in CD4⫹CD25⫹ T cells in the me/me mice is directly due to the lack of SHP-1 during thymic development, and not a secondary effect in the context of the mouse. Selective enrichment of CD4⫹CD25⫹ T cells upon exposure of cognate Ag The above data suggested a model that stronger TCR-mediated signaling, due to the absence of the negative regulator SHP-1, favors the generation of CD4⫹CD25⫹ Treg cells. A logical extension of this hypothesis is that, even in wild-type mice, the percentage of Treg cells should be increased when cognate Ag is present during the selection process. Consistent with this hypothesis, transgenic expression of the cognate Ag in the thymus of TCR-Tg mice has been shown to promote the development of Treg cells (11–13);

however, the design of these in vivo studies did not allow for a careful dose response of the cognate peptide. To directly determine whether a cognate peptide can enhance the appearance of CD4⫹CD25⫹ T cells and to determine the effect of various peptide concentrations in this process, we set up FTOCs from ⫹/⫹

FIGURE 4. CD4⫹CD25⫹ Treg cells from me/me are as effective in suppression of proliferation as ⫹/⫹ CD4⫹CD25⫹ Treg cells. a, SHP-1 was immunoprecipitated from cleared lysates of the indicated number of electronically purified splenic CD4⫹CD25⫹ or CD4⫹CD25⫺ T cells. Immunoprecipitates were resolved by 10% SDS-PAGE and subjected to antiSHP-1 immunoblotting. b, Splenic CD4⫹CD25⫹ and CD4⫹CD25⫺ cells were electronically purified from ⫹/⫹ and me/me DO11.10 TCR-Tg mice. ⫹/⫹ or me/me CD4⫹CD25⫹ T cells (Treg) were added to 2.5 ⫻ 104 of ⫹/⫹ CD4⫹CD25⫺ T cells (Teff) at the indicated ratios along with 5 ⫻ 104 APC and CD3 Ab (145-2c11; 6 ␮g/ml). Proliferation of ⫹/⫹ CD4⫹CD25⫺ T cells is set as 100%. Error bars denote ⫾SEM. Results shown are representative of three independent experiments.


6633

FIGURE 5. CD4⫹CD25⫹ T cells are thymus derived. a and b, Thymocytes and splenocytes from 4-day-old ⫹/⫹ and me/me BALB/c (a) and DO11.10 TCR-Tg (b) mice were isolated and stained with the indicated Abs. A total of 150,000 live-gated thymocytes or splenocytes was analyzed per point. Indicated subpopulations (based on CD4 and CD8 expression) were further analyzed for CD25 expression. Percentages of CD25⫹ within the CD4⫹ subpopulation are shown. c, Thymi were removed from embryos at day 15 of gestation (day 0) and cultured in Transwell plates. On day 7 of culture, the thymi were removed from culture, and the thymocytes were isolated and stained for flow cytometry with CD4-PerCP, CD8-PE, and CD25-allophycocyanin. Approximately 105 live-gated thymocytes were collected for each point. Profiles of CD8 vs CD4 are shown. CD4⫹ SP thymocytes were further analyzed for CD25 expression. Percentages of CD25⫹ cells within the CD4 SP population are indicated.

DO11.10 mice and cocultured them with increasing concentrations of OVA peptide (Fig. 6). We made three notable observations. First, we observed an increase in the percentage of CD4⫹CD25⫹ cells within the CD4⫹ SP subpopulation at peptide concentrations coincident with negative selection, as manifested by a loss of the DP population (⬎1 ␮g/ml OVA peptide; Fig. 6). Second, the absolute cell number of CD4⫹CD25⫹ T cells remained relatively constant at all concentrations of OVA peptide, indicating that increasing the peptide concentration per se (in turn, greater strength of signal) did not change the absolute number of CD4⫹CD25⫹ T cells that emerge under these conditions. Third, even at peptide concentrations ⬎10 times required for negative selection, no decrease in numbers of the CD4⫹CD25⫹ T cell population was detectable, suggesting a selective resistance of the Treg cell population to conditions of negative selection. Concurrent analysis of V␣2 usage in the CD4⫹CD25⫹ and CD4⫹CD25⫺ subpopulations demonstrates that, at least at the lower concentrations of peptide, the percentage of V␣2-positive cells are relatively constant (data not shown). This observation indicated that the resulting CD4⫹CD25⫹ cells are Treg cells and not activated T cells. The above data finding a constant number of CD4⫹CD25⫹ T cells under different peptide concen-

trations suggests that a fraction of the developing T cells in the thymus are precommitted to become CD4⫹CD25⫹ cells, and that they can withstand an environment/signals strong enough to cause negative selection of non-Treg cells (see Discussion). The relatively constant number of CD4⫹CD25⫹ cells observed in FTOCs is different from what we observed in the whole organism, where the motheaten mice, especially in the DO11.10 background, showed an increase in the absolute number of thymic and splenic CD4⫹CD25⫹ Treg cells. This can be readily reconciled by the fact that no new prothymocytes enter the FTOCs and that the thymus cannot be replenished during changes in thymocyte composition. In contrast, new progenitor cells can replenish the thymus in the mouse, resulting in a higher throughput during states of increased precursor influx, thereby leading to increased accumulation of CD4⫹CD25⫹ Treg cells. Expanding the above study on CD4⫹CD25⫹ T cell development in ⫹/⫹:DO11.10 FTOCs, we tested the effect of combining SHP-1 deficiency and addition of cognate peptide during thymic development. We hypothesized that me/me FTOCs should be more sensitive to OVA peptide and that the amount of peptide needed to promote the generation of CD4⫹CD25⫹ Treg cells would be decreased in the absence of SHP-1. FTOCs from ⫹/⫹ and me/me

6634


FIGURE 6. Addition of increasing amounts of OVA peptide to FTOCs results in increased percentages of CD4⫹CD25⫹ T cells. Thymi were removed from embryos at day 15 of gestation (day 0) and cultured in Transwell plates with the aboveindicated amounts of OVA peptide on days 2–5. On day 7 of culture, the thymi were removed from culture, and the thymocytes were isolated and stained for flow cytometry with CD4PerCP, CD8-allophycocyanin, and CD25-PE. Approximately 105 livegated thymocytes were collected for each point. CD8 vs CD4 profiles are shown for each OVA concentration (top panels). CD25 surface expression on CD4⫹ SP thymocytes is shown in the bottom panels. Percentages of CD25⫹ cells within the CD4⫹ SP subpopulation are indicated. The experiments shown are representative of at least four thymic lobes per concentration.

DO11.10 mice were cocultured with OVA peptide and analyzed for the presence of CD4⫹CD25⫹ cells. At low concentrations of OVA peptide (0.5 or 1 ␮g), the increase in the percentage of CD4⫹CD25⫹ cells within the CD4⫹ SP population of ⫹/⫹ FTOCs was absent or minimal. However, me/me FTOCs contained a readily detectable increase in the percentage of CD4⫹CD25⫹ cells in response to the same OVA peptide concentration (Fig. 7). The enrichment of CD4⫹CD25⫹ T cells in the me/me FTOCs was concurrent with extensive deletion. Taken together, our data suggest that the CD4⫹CD25⫹ Treg cells are precommitted to the Treg cell lineage before their encounter with the negatively selecting thymic environment. Our findings also highlight the importance of SHP-1-dependent regulation of TCR signal strength as a contributing factor in determining the Treg vs non-Treg cell ratios.

Discussion We and others have previously reported that the tyrosine phosphatase SHP-1 plays a fundamentally important role in regulating TCR signal strength (28, 29). This translates to SHP-1 setting thresholds for positive and negative selection during thymic T cell development (31–34). The presence of high-affinity peptides during thymic selection has been shown to promote Treg cell development (11, 12, 14). It was proposed that this was directly related to heightened signaling via the TCR during the selection process. In this study, we took two different approaches to test this model and to address whether modulation of specific signaling molecules can promote increased Treg development. We analyzed mice lacking SHP-1 for the presence of Treg cells and observed an increase in the percentage of Treg cells in the me/me mice compared with control mice. Based on this finding, we hypothesized that controlled modulation of TCR signal strength through varying peptide concentrations would also affect Treg development. Using a FTOC model system where we added increasing concentrations of cognate peptide, we have presented evidence that Treg cell develop-

ment is favored under conditions of high signal strength that promote negative selection. Several lines of evidence confirmed that the CD4⫹CD25⫹ T cells present in me/me mice were functional Treg cells. These include the following: their anergic phenotype when stimulated in vitro, their ability to suppress the proliferation of Teff cells, their surface marker expression, the presence of Foxp3 mRNA, and the coexpression of a second endogenous TCR␣ chain in the case of DO11.10 TCR-Tg background. Moreover, these CD4⫹CD25⫹ T cells seen in the me/me mice were thymus derived as confirmed using the FTOC system. Our analyses also revealed a relative as well as an absolute increase in numbers of Treg cells in spleens of me/me mice. This is most likely a direct result of the increased output of Treg cells from the thymus, although an additional peripheral expansion is still possible. There is some evidence for Treg cell development in the periphery, although the precise mechanisms are unclear (20, 40, 51). Although all of the spleens analyzed were from 14- to 17day-old mice due to the limited life span of the me/me mice, environment-driven alterations in Treg numbers in the periphery could not be excluded. Whether defects in other cell types in me/me mice influence peripheral Treg numbers remains to be established. Additional studies are needed before making definitive conclusions about thymus-independent Treg cell development in me/me mice. The increased percentage of thymic CD4⫹CD25⫹ Treg cells in the me/me background suggested that absence of SHP-1 favored Treg cell development over effector CD4⫹CD25⫺ T cells. There are at least two potential, not mutually exclusive, models that can explain this phenotype (Fig. 8). Increased-selection model In this model, thymic selection of Treg cells would be limited by a threshold that has to be overcome, similar to the classical positiveand negative-selection criteria. This threshold would, at least in


6635

FIGURE 7. FTOCs derived from me/me embryos are more sensitive to OVA peptide than FTOCs derived from ⫹/⫹ littermates. Thymi were removed from embryos at day 15 of gestation (day 0). The two lobes of each thymus were separated, and one was cultured without OVA peptide, whereas the other was cultured with the indicated amounts of OVA peptide on days 2–5. On day 7 of culture, thymi were removed from culture and stained with CD4-PerCP, CD8-allophycocyanin, and CD25-PE. Approximately 105 live-gated thymocytes were collected for each point. CD8 vs CD4 profiles are shown for each OVA concentration (top panels). CD4⫹ SP thymocytes were further analyzed for CD25 surface expression (bottom panels). Percentages of CD25⫹ cells within the CD4⫹ SP subpopulations are indicated. The data shown are derived from matched pairs of thymic lobes isolated from the same thymus and are representative of at least 13 thymi for each concentration. The data for FTOCs exposed to 500 and 1000 ng OVA peptide, respectively, were derived from different experiments, which required setting the gates differently between the experiments. However within each concentration, ⫹/⫹ and me/me FTOCs were derived from littermates.

part, be regulated by SHP-1. In other words, the positive selection of Treg cells would be enhanced by the absence of SHP-1. However, because the loss of SHP-1 would have lowered the threshold for Treg and non-Treg cells, yet the me/me mice are selectively enriched for Treg cells, the model would require that Treg cells be more affected from lowering the threshold than the non-Treg cells. Precommitment/selection model In this model, a set fraction of thymocytes would be precommitted to go into the Treg lineage before they undergo the regular selection process. This could either be achieved by a stochastic event before the selection process or through an encounter of the developing thymocyte with a special cell type that provides an instructional signal. Cells thereby committed to the Treg lineage would still require a strong signal to be positively selected but would be less susceptible to negative selection. In the context of this model, two other previous observations are of interest. First, conditions where one can expect greater negative selection coincide with increased appearance of Treg cells, suggesting that Treg cells are less susceptible to negative selection (11–13). Second, the higher TCR signal strength in the absence of SHP-1 leads to enhanced negative selection (31). Therefore, in the me/me background, where there is greater negative selection of non-Treg cells, this could lead to a selective enrichment of precommitted Treg cells. If model 1 were true, conditions of heightened TCR signal strength would lead to an increase in the absolute number of thymic Treg cells. If model 2 were correct, the absolute number of Treg

cells, which is dictated by the fraction of cells precommitted to this lineage, would not change, unless there is also an increased expansion of the Treg population; however, the ratio of Treg cells vs non-Treg cells would be altered. Therefore, the absence of a change in the absolute number of Treg cells supports model 2, whereas an increase in the absolute cell number would require further studies to distinguish between the two models. Although we did see an increase in the absolute number of thymic Treg cells in the me/me DO11.10 TCR-Tg mice, the interpretation of this result was complicated by a potentially elevated influx of prothymocytes as a consequence of changes in the thymocyte subpopulations (31). A recent study addressed the mechanism of controlling progenitor recruitment/precursor influx into the thymus (52). The authors found that the size of the double-negative (DN) pool and specifically the cell number at the DN3 stage control the influx of new progenitors. In this context, it is of particular interest that thymi from DO11.10 me/me mice, which have greatly reduced numbers of DN3 cells (data not shown) and therefore potentially the greatest influx of thymocyte precursor cells have the greatest increase in the number of CD4⫹CD25⫹ Treg cells. In contrast, non-Tg me/me mice do not show such a great decrease in the DN3 population and do not show a strong increase in the absolute number of thymic CD4⫹CD25⫹ Treg cells. Increased influx would have resulted in increased absolute numbers independent of which model was operational. To circumvent the complication of changes in influx, we used the FTOC model system that precludes new influx of progenitors into the thymus. An additional advantage of the FTOC

6636


FIGURE 8. Models for thymic CD4⫹CD25⫹ Treg cell development. In the increased-selection model (model 1), thymocytes at the DP stage are undecided and can develop into Treg and non-Treg cells. CD4⫹CD25⫹ Treg cells are selected in response to a defined range of TCR signaling strength that falls between positive selection of non-Treg cells and negative selection/deletion (left panel). Under conditions of increased TCR signaling strength (right panel), such as exposure of the TCR to its cognate peptide or loss of the negative regulator SHP-1, the selection process shifts toward positive selection of CD4⫹CD25⫹ Treg cells and greater deletion of non-Treg cells, resulting in an increase in the absolute number of CD4⫹CD25⫹ T cells. In contrast, in the precommitment/selection model (model 2), a defined fraction of the CD4⫹CD8⫹ DP thymocytes is already committed to the CD4⫹CD25⫹ Treg cell lineage. The remaining DP cells have the potential to become non-Treg cells upon passing of positive/negative selection, but lost the potential to develop into Treg cells. The cells precommitted to the Treg lineage might still require further signaling and undergo an additional selection process to acquire the Treg phenotype, but their absolute numbers are unaffected by the TCR signal strength as long as it is above a certain threshold. Moreover, even when exposed to high TCR-mediated signaling, these developing CD4⫹CD25⫹ T cells are resistant to deletion. Therefore under conditions of increased TCR signaling strength (right panel), there is a increase in the percentage of CD4⫹CD25⫹ T cells due to the selective loss of CD4⫹CD25⫺ non-Treg cells via increased negative selection.

system is that it allows the addition of a selecting peptide at controlled concentrations, enabling us to correlate signal strength with the development/enrichment of Treg cells. When FTOCs (derived from ⫹/⫹ DO11.10 TCR-Tg mice) were treated with increasing amounts of OVA peptide, the absolute numbers of CD4⫹CD25⫹ cells stayed relatively constant throughout the various OVA concentrations, despite the relative increase within the CD4⫹ population. This argues against model 1, because this should have led to higher absolute numbers of CD4⫹CD25⫹ T cells. However, the data are consistent with the precommitment/ selection model (model 2), where a set fraction of developing thymocytes are committed to the Treg lineage, yet these cells are less susceptible to negative selection. Moreover, we observed that the CD4⫹CD25⫹ T cell population recovered after OVA peptide treatment was still enriched for V␣2 compared with the CD4⫹CD25⫺ T cell population. This is consistent with the CD4⫹CD25⫹ T cell population being precommitted to the Treg lineage and not representing CD4⫹CD25⫺ thymocytes newly recruited to this lineage. The selective loss of non-Treg lineage cells at deletion-inducing peptide concentrations would lead to a relative increase in Treg cells, without a concomitant increase in absolute numbers. An important question is at what stage of thymic T cell development do cells commit to the Treg cell lineage. Based on our present knowledge, commitment could occur before any selection event or somewhere during the developmental process guided by extrinsic cues; alternatively, Treg cells could also arise from a different precursor pool. Careful future studies are necessary to address this question.

It is interesting that the cells committed to the Treg lineage are very resistant to negative selection in the FTOC model system. In fact, we failed to detect deletion of the CD4⫹CD25⫹ population even at ⬎10 times the concentration required to achieve negative selection of CD4⫹CD25⫺ T cells. The CD4⫹CD25⫹ T cells that survived were still KJ1-26 positive, and therefore should have been capable of recognizing OVA peptide. However, it is possible that the CD4⫹CD25⫹ T cells were less sensitive to OVA peptide due to their coexpression of an endogenous TCR␣ chain. Resulting changes in the expression or signaling of the clonotypic TCR on this T cell population compared with the CD4⫹CD25⫺ T cells might have rendered the CD4⫹CD25⫹ T cells more resistant to deletion. Future studies are needed to better understand the contribution of the endogenous TCR␣ chain to the development and function of TCR-Tg CD4⫹CD25⫹ Treg cells. We also considered the possibility that, in the FTOCs, the OVA presentation occurred through thymic subcompartments that do not support Treg deletion; however, this seems unlikely because the me/me mice also show a selective enrichment of the CD4⫹CD25⫹ T cell population. To further determine the relationship between negative selection and the emergence of Treg cells vs non-Treg cells, we combined the genetic approach of using me/me mice and the addition of peptides in the FTOC system. In me/me FTOCs, where the negative deletion of non-Treg cells occurs at lower OVA peptide concentrations, the relative increase in Treg cells was concurrently detected. This further supported our hypothesis that TCR signal strength, and its regulation by SHP-1, influence the ratio of Treg and non-Treg cells.

The Journal of Immunology In this regard, it is interesting that pharmacological interference with the glucocorticoid/glucocorticoid receptor pathway, thought to lower the threshold for T cell activation, also influences the fraction of Treg/non-Treg cell ratios (53). Although most of the previous models of Treg cell development suggest that CD4⫹CD25⫹ Treg cells are generated in direct response to increased signaling via the TCR, our data support an alternative model where Treg cells are precommitted to this lineage before encountering signals that cause negative deletion, and that Treg cells are relatively resistant to negative selection. At the same time, developing CD4⫹CD25⫺ non-Treg thymocytes are sensitive to strong signaling events, which trigger their deletion. Therefore, under conditions of high signal strength, the survival/development of Treg cells is favored compared with regular CD4⫹CD25⫺ Teff cells. Our data do not preclude that Treg cells still need a higher affinity peptide-TCR interaction to be positively selected as it is postulated in the existing models of Treg cell development. Although the precommitment/selection model proposed here represents one way to generate natural CD4⫹CD25⫹ Treg cells, it does not exclude other means of generating Treg cells, such as extrathymic expansion or new recruitment. The precommitment/selection model is consistent with previous data where an increase in CD4⫹CD25⫹ Treg cells was observed under conditions of exposure to agonist self-peptide (11, 13, 14, 54). Jordan et al. (11) showed that coexpression of a Tg TCR with its cognate peptide results in loss or decrease of the CD4⫹CD25⫺ T cell population, whereas there is an increase in CD4⫹CD25⫹ Treg cells at the same time. In contrast, expression of a low-affinity TCR with its cognate peptide allows the generation of CD4⫹CD25⫺ with only a small percentage of the T cells being CD4⫹CD25⫹ Treg cells. In a recent report, this study was extended by demonstrating that radioresistant stromal cells, expressing the cognate peptide, are capable of mediating both deletion as well as selection of CD4⫹CD25⫹ Treg cells. Moreover, the authors show that populations of CD4⫹CD25⫹ Treg cells and clonally related CD4⫹CD25⫺ T cells are generated in the same mouse (54). This is in agreement with our model where conditions of heightened deletion result in an increase in CD4⫹CD25⫹ Treg cells, and it is consistent with the hypothesis that a selective enrichment of Treg cells is not directly resulting from heightened TCR signaling during the selection process of the Treg cells but rather an indirect effect due to selective deletion of non-Treg cells and a relative deletion resistance of the Treg cell population. Similarly, Kawahata et al. (13) demonstrate that transgenic expression of OVA peptide in DO11.10 TCR-Tg mice results in clonal deletion of DO11.10 T cells simultaneously with an increase in the number and percentage of CD4⫹CD25⫹ T cells, indicating that positive selection of CD4⫹CD25⫹ Treg cells occurs in parallel to negative selection of CD4⫹CD25⫺ T cells. These findings are comparable to what we observed in the me/me DO11.10 mice, whereas in the DO11.10derived FTOCs upon exposure to peptide, we observed concurrent deletion of non-Treg cells and positive selection of Treg cells without an increase in the absolute number of CD4⫹CD25⫹ T cells. As was discussed above, in mice where there is an ongoing influx of T cell precursors into the thymus, an increase in the absolute number of CD4⫹CD25⫹ Treg cells is still consistent with our precommitment/selection model. Interestingly, while we were writing this paper, van Santen et al. (55) reported a finding similar to ours. In an elegant study using triple transgenic mice, they showed that graded expression of the cognate peptide caused increased appearance of CD4⫹CD25⫹ Treg cells. Further analysis demonstrated that the number of CD4⫹CD25⫹ T cells increased as a result of a selective survival of this population and not due to induced differentiation.

6637 Taken together, these data suggest that CD4⫹C25⫹ Treg cells are selectively enriched under conditions of increased negative selection. Our observations that the percentage of Treg cells can be altered through the modulation of TCR signal strength, either by specific signaling molecules such as SHP-1, or by changing the concentration of selecting peptides, provide an exciting avenue for the manipulation of Treg numbers in vivo. Given the link between Treg cells and autoimmunity, the above studies may have important implications for physiology and pathophysiology of the immune system.

Acknowledgments Unless otherwise stated, J. D. Carter performed all the experiments in this study with the technical support of G. M. Calabrese. M. Naganuma in Dr. Peter Ernst’s laboratory executed the Foxp3 PCR, and we thank Dr. Ernst for his contribution. We are grateful to Drs. Tim Bender, Patrick Lyons, Kodi Ravichandran, and Ken Tung for critical reading of the manuscript and their comments and suggestions.

Disclosures The authors have no financial conflict of interest.

References 1. Sprent, J., and H. Kishimoto. 2002. The thymus and negative selection. Immunol. Rev. 185: 126 –135. 2. Gebe, J. A., B. A. Falk, K. A. Rock, S. A. Kochik, A. K. Heninger, H. Reijonen, W. W. Kwok, and G. T. Nepom. 2003. Low-avidity recognition by CD4⫹ T cells directed to self-antigens. Eur. J. Immunol. 33: 1409 –1417. 3. Arnold, B. 2002. Levels of peripheral T cell tolerance. Transpl. Immunol. 10: 109 –114. 4. Shevach, E. M. 2002. CD4⫹CD25⫹ suppressor T cells: more questions than answers. Nat. Rev. Immunol. 2: 389 – 400. 5. Nishizuka, Y., and T. Sakakura. 1969. Thymus and reproduction: sex-linked dysgenesia of the gonad after neonatal thymectomy in mice. Science 166: 753–755. 6. Gershon, R. K., and K. Kondo. 1971. Infectious immunological tolerance. Immunology 21: 903–914. 7. Annacker, O., R. Pimenta-Araujo, O. Burlen-Defranoux, and A. Bandeira. 2001. On the ontogeny and physiology of regulatory T cells. Immunol. Rev. 182: 5–17. 8. Suri-Payer, E., A. Z. Amar, A. M. Thornton, and E. M. Shevach. 1998. CD4⫹CD25⫹ T cells inhibit both the induction and effector function of autoreactive T cells and represent a unique lineage of immunoregulatory cells. J. Immunol. 160: 1212–1218. 9. Thornton, A. M., and E. M. Shevach. 1998. CD4⫹CD25⫹ immunoregulatory T cells suppress polyclonal T cell activation in vitro by inhibiting interleukin 2 production. J. Exp. Med. 188: 287–296. 10. Taylor, P. A., C. J. Lees, and B. R. Blazar. 2002. The infusion of ex vivo activated and expanded CD4⫹CD25⫹ immune regulatory cells inhibits graft-versus-host disease lethality. Blood 99: 3493–3499. 11. Jordan, M. S., A. Boesteanu, A. J. Reed, A. L. Petrone, A. E. Holenbeck, M. A. Lerman, A. Naji, and A. J. Caton. 2001. Thymic selection of CD4⫹CD25⫹ regulatory T cells induced by an agonist self-peptide. Nat. Immunol. 2: 301–306. 12. Walker, L. S., A. Chodos, M. Eggena, H. Dooms, and A. K. Abbas. 2003. Antigen-dependent proliferation of CD4⫹CD25⫹ regulatory T cells in vivo. J. Exp. Med. 198: 249 –258. 13. Kawahata, K., Y. Misaki, M. Yamauchi, S. Tsunekawa, K. Setoguchi, J. Miyazaki, and K. Yamamoto. 2002. Generation of CD4⫹CD25⫹ regulatory T cells from autoreactive T cells simultaneously with their negative selection in the thymus and from nonautoreactive T cells by endogenous TCR expression. J. Immunol. 168: 4399 – 4405. 14. Apostolou, I., A. Sarukhan, L. Klein, and H. von Boehmer. 2002. Origin of regulatory T cells with known specificity for antigen. Nat. Immunol. 3: 756 –763. 15. Bensinger, S. J., A. Bandeira, M. S. Jordan, A. J. Caton, and T. M. Laufer. 2001. Major histocompatibility complex class II-positive cortical epithelium mediates the selection of CD4⫹25⫹ immunoregulatory T cells. J. Exp. Med. 194: 427– 438. 16. Baecher-Allan, C., V. Viglietta, and D. A. Hafler. 2002. Inhibition of human CD4⫹CD25⫹high regulatory T cell function. J. Immunol. 169: 6210 – 6217. 17. Klein, L., K. Khazaie, and H. von Boehmer. 2003. In vivo dynamics of antigenspecific regulatory T cells not predicted from behavior in vitro. Proc. Natl. Acad. Sci. USA 100: 8886 – 8891. 18. Thornton, A. M., and E. M. Shevach. 2000. Suppressor effector function of CD4⫹CD25⫹ immunoregulatory T cells is antigen nonspecific. J. Immunol. 164: 183–190. 19. Yamagiwa, S., J. D. Gray, S. Hashimoto, and D. A. Horwitz. 2001. A role for TGF-␤ in the generation and expansion of CD4⫹CD25⫹ regulatory T cells from human peripheral blood. J. Immunol. 166: 7282–7289. 20. Kuniyasu, Y., T. Takahashi, M. Itoh, J. Shimizu, G. Toda, and S. Sakaguchi. 2000. Naturally anergic and suppressive CD25⫹CD4⫹ T cells as a functionally and phenotypically distinct immunoregulatory T cell subpopulation. Int. Immunol. 12: 1145–1155.

6638 21. Plutzky, J., B. G. Neel, and R. D. Rosenberg. 1992. Isolation of a novel SRC homology 2 (SH2) containing tyrosine phosphatase. Proc. Natl. Acad. Sci. USA 89: 1123–1127. 22. Matthews, R. J., D. B. Bowne, E. Flores, and M. L. Thomas. 1992. Characterization of hematopoietic intracellular protein tyrosine phosphatases: description of a phosphatase containing an SH2 domain and another enriched in proline-, glutamic acid-, serine-, and threonine rich sequences. Mol. Cell. Biol. 12: 2396 –2405. 23. Yi, T., J. L. Cleveland, and J. N. Ihle. 1992. Protein tyrosine phosphatase containing SH2 domains: characterization, preferential expression in hematopoietic cells, and localization to human chromosome 12p12–13. Mol. Cell. Biol. 12: 836 – 846. 24. Shen, S.-H., L. Bastien, B. Posner, and P. Chretien. 1991. A protein-tyrosine phosphatase with sequence similarity to the SH2 domain of the protein-tyrosine kinases. Nature 352: 736 –739. 25. Neel, B. G., J. Plutzky, R. D. Rosenberg, R. M. Freeman, Jr., U. Lorenz, and R. J. Lechleider. 1993. Molecular cloning and characterization of a subfamily of non-transmembrane PTPs containing src-homology 2 domains. Adv. Prot. Phosphatases 7: 127–151. 26. Shultz, L. D., P. A. Schweitzer, T. V. Rajan, T. Yi, J. N. Ihle, R. J. Matthews, M. L. Thomas, and D. R. Beier. 1993. Mutations at the murine motheaten locus are within the hematopoietic cell protein phosphatase (HCPH) gene. Cell 73: 1445–1454. 27. Tsui, H. W., K. A. Siminovitch, L. deSouza, and F. W. L. Tsui. 1993. Motheaten and viable motheaten mice have mutations in the haematopoietic cell phosphatase gene. Nat. Genet. 4: 124 –129. 28. Lorenz, U., K. S. Ravichandran, S. J. Burakoff, and B. G. Neel. 1996. Lack of SHPTP1 results in src-family kinase hyperactivation and thymocyte hyperresponsiveness. Proc. Natl. Acad. Sci. USA 93: 9624 –9629. 29. Pani, G., K.-D. Fischer, I. Mlinaric-Rascan, and K. A. Siminovitch. 1996. Signaling capacity of the T cell antigen receptor is negatively regulated by the PTP1c tyrosine phosphatase. J. Exp. Med. 184: 839 – 852. 30. Murphy, K. M., A. B. Heimberger, and D. Y. Loh. 1990. Induction by antigen of intrathymic apoptosis of CD4⫹CD8⫹ TCRlo thymocytes in vivo. Science 250: 1720 –1723. 31. Carter, J. D., B. G. Neel, and U. Lorenz. 1999. The tyrosine phosphatase SHP-1 influences thymocyte selection by setting TCR signaling thresholds. Int. Immunol. 11: 1999 –2014. 32. Zhang, J., A. K. Somani, D. Yuen, Y. Yang, P. E. Love, and K. A. Siminovitch. 1999. Involvement of the SHP-1 tyrosine phosphatase in regulation of T cell selection. J. Immunol. 163: 3012–3021. 33. Johnson, K. G., F. G. LeRoy, L. K. Borysiewicz, and R. J. Matthews. 1999. TCR signaling thresholds regulating T cell development and activation are dependent upon SHP-1. J. Immunol. 162: 3802–3813. 34. Plas, D. R., C. B. Willians, G. J. Kersh, L. S. White, J. M. White, S. Paust, T. Ulyanova, P. M. Allen, and M. L. Thomas. 1999. The tyrosine phosphatase SHP-1 regulates thymocyte positive selection. J. Immunol. 162: 5680 –5684. 35. Pacholczyk, R., P. Kraj, and L. Ignatowicz. 2002. Peptide specificity of thymic selection of CD4⫹CD25⫹ T cells. J. Immunol. 168: 613– 620. 36. Denning, T. L., H. Qi, R. Konig, K. G. Scott, M. Naganuma, and P. B. Ernst. 2003. CD4⫹ Th cells resembling regulatory T cells that inhibit chronic colitis differentiate in the absence of interactions between CD4 and class II MHC. J. Immunol. 171: 2279 –2286. 37. DeLuca, D., J. A. Bluestone, L. D. Shultz, S. O. Sharrow, and Y. Tatsumi. 1995. Programmed differentiation of murine thymocytes during fetal thymus organ culture. J. Immunol. Methods 178: 13–29.

DEFICIENCY OF SHP-1 CAUSES ENRICHMENT OF Treg CELLS 38. Swat, W., M. Dessing, H. von Boehmer, and P. Kisielow. 1993. CD69 expression during selection and maturation of CD4⫹8⫹ thymocytes. Eur. J. Immunol. 23: 739 –746. 39. Merkenschlager, M., D. Graf, M. Lovatt, U. Bommhardt, R. Zamoyska, and A. G. Fisher. 1997. How many thymocytes audition for selection? J. Exp. Med. 186: 1149 –1158. 40. Suto, A., H. Nakajima, K. Ikeda, S. Kubo, T. Nakayama, M. Taniguchi, Y. Saito, and I. Iwamoto. 2002. CD4⫹CD25⫹ T-cell development is regulated by at least 2 distinct mechanisms. Blood 99: 555–560. 41. Read, S., S. Mauze, C. Asseman, A. Bean, R. Coffman, and F. Powrie. 1998. CD38⫹CD45RBlowCD4⫹ T cells: a population of T cells with immune regulatory activities in vitro. Eur. J. Immunol. 28: 3435–3447. 42. Read, S., V. Malmstrom, and F. Powrie. 2000. Cytotoxic T lymphocyte-associated antigen 4 plays an essential role in the function of CD25⫹CD4⫹ regulatory cells that control intestinal inflammation. J. Exp. Med. 192: 295–302. 43. Shimizu, J., S. Yamazaki, T. Takahashi, Y. Ishida, and S. Sakaguchi. 2002. Stimulation of CD25⫹CD4⫹ regulatory T cells through GITR breaks immunological self-tolerance. Nat. Immunol. 3: 135–142. 44. McHugh, R. S., M. J. Whitters, C. A. Piccirillo, D. A. Young, E. M. Shevach, M. Collins, and M. C. Byrne. 2002. CD4⫹CD25⫹ immunoregulatory T cells: gene expression analysis reveals a functional role for the glucocorticoid-induced TNF receptor. Immunity 16: 311–323. 45. Caramalho, I., T. Lopes-Carvalho, D. Ostler, S. Zelenay, M. Haury, and J. Demengeot. 2003. Regulatory T cells selectively express Toll-like receptors and are activated by lipopolysaccharide. J. Exp. Med. 197: 403– 411. 46. Itoh, M., T. Takahashi, N. Sakaguchi, Y. Kuniyasu, J. Shimizu, F. Otsuka, and S. Sakaguchi. 1999. Thymus and autoimmunity: production of CD25⫹CD4⫹ naturally anergic and suppressive T cells as a key function of the thymus in maintaining immunologic self-tolerance. J. Immunol. 162: 5317–5326. 47. Fontenot, J. D., M. A. Gavin, and A. Y. Rudensky. 2003. Foxp3 programs the development and function of CD4⫹CD25⫹ regulatory T cells. Nat. Immunol. 4: 330 –336. 48. Hori, S., T. Nomura, and S. Sakaguchi. 2003. Control of regulatory T cell development by the transcription factor Foxp3. Science 299: 1057–1061. 49. Khattri, R., T. Cox, S. A. Yasayko, and F. Ramsdell. 2003. An essential role for Scurfin in CD4⫹CD25⫹ T regulatory cells. Nat. Immunol. 4: 337–342. 50. Asano, M., M. Toda, N. Sakaguchi, and S. Sakaguchi. 1996. Autoimmune disease as a consequence of developmental abnormality of a T cell subpopulation. J. Exp. Med. 184: 387–396. 51. Thorstenson, K. M., and A. Khoruts. 2001. Generation of anergic and potentially immunoregulatory CD25⫹CD4 T cells in vivo after induction of peripheral tolerance with intravenous or oral antigen. J. Immunol. 167: 188 –195. 52. Prockop, S. E., and H. T. Petrie. 2004. Regulation of thymus size by competition for stromal niches among early T cell progenitors. J. Immunol. 173: 1604 –1611. 53. Stephens, G. L., and L. Ignatowicz. 2003. Decreasing the threshold for thymocyte activation biases CD4⫹ T cells toward a regulatory (CD4⫹CD25⫹) lineage. Eur. J. Immunol. 33: 1282–1291. 54. Lerman, M. A., J. Larkin III, C. Cozzo, M. S. Jordan, and A. J. Caton. 2004. CD4⫹CD25⫹ regulatory T cell repertoire formation in response to varying expression of a neo-self-antigen. J. Immunol. 173: 236 –244. 55. van Santen, H. M., C. Benoist, and D. Mathis. 2004. Number of T reg cells that differentiate does not increase upon encounter of agonist ligand on thymic epithelial cells. J. Exp. Med. 200: 1221–1230.