RasGRP1 and Sos are two Ras-guanyl-nucleotide exchange factors that link TCR signal transduction to Ras and MAPK activation. Recent studies demonstrate positive selection of developing thymocytes is crucially dependent on RasGRP1, whereas negative selection of autoreactive thymocytes appears to be RasGRP1 independent. However, the role of RasGRP1 in T regulatory (Treg) cell development and function is unknown. In this study, we characterized the development and function of CD4+CD25+Foxp3+ and CD8+CD44highCD122+ Treg lineages in RasGRP1−/− mice. Despite impaired CD4 Treg cell development in the thymus, the periphery of RasGRP1−/− mice contained significantly increased frequencies of CD4+Foxp3+ Treg cells that possessed a more activated cell surface phenotype. Furthermore, on a per cell basis, CD4+Foxp3+ Treg cells from mutant mice are more suppressive than their wild-type counterparts. Our data also suggest that the lymphopenic environment in the mutant mice plays a dominant role of favored peripheral development of CD4 Treg cells. These studies suggest that whereas RasGRP1 is crucial for the intrathymic development of CD4 Treg cells, it is not required for their peripheral expansion and function. By contrast to CD4+CD25+Foxp3+ T cells, intrathymic development of CD8+CD44highCD122+ Treg cells is unaffected by the RasGRP1−/− mutation. Moreover, RasGRP1−/− mice contained greater numbers of CD8+CD44highCD122+ T cells in the spleen, relative to wild-type mice. Activated CD8 Treg cells from RasGRP1−/− mice retained their ability to synthesize IL-10 and suppress the proliferation of wild-type CD8+CD122− T cells, albeit at a much lower efficiency than wild-type CD8 Treg cells.
During the process of thymic education, only thymocytes expressing TCRs capable of distinguishing between self and nonself are selected for survival and differentiate into mature T cells. According to the strength of signal hypothesis, thymocytes expressing TCRs with weak or modest affinity for self Ags undergo positive selection and become mature T cells. By contrast, developing T cells expressing high affinity receptors for self Ags are clonally deleted, a process referred to as negative selection. However, negative selection is imperfect because autoreactive lymphocytes escape clonal deletion and can be detected within the peripheral lymphocyte compartment (1, 2). To keep these autoreactive T cell clones in check, a dedicated lineage of T regulatory (Treg)4 cells functions as an important fail-safe mechanism to prevent catastrophic effect of unchecked immune responses (3, 4). Three main cell types have been considered as potential Treg cell subsets, as follows: CD4+CD25+ Treg cells (5, 6), CD8αα intestinal epithelial lymphocytes (7), and NKT cells (3). All are thought to be induced by high affinity interactions between self peptide:MHC with the TCR on developing T cells in the thymus (1, 8). In addition to these three subsets, recent studies indicate normal mice also possess a population of CD8+CD44highCD122+ T cells that function as regulatory cells and can perform roles distinct from CD4 Treg cells in suppressing T cell activation (9, 10, 11, 12).
Of all the lineages of Treg cells, the CD4+CD25+ subset has been the most extensively characterized. CD4+CD25+ T cells develop naturally in normal individuals and are readily detectable in the thymus and secondary lymphoid organs in mice, rats, and humans, where they make up ∼2–10% of the total CD4+ cells (13). In addition to CD25, CD4 Treg cells also express high levels of CTLA-4 and glucocorticoid-induced tumor necrosis factor receptor-related protein (14). However, the most distinguishing feature of CD4 Treg cells is their expression of Foxp3, a member of the forkhead family of transcription factor (8). The expression of Foxp3 is both necessary and sufficient for the development and function of CD4 Treg cells (5, 15). Additionally, studies have found the TCR repertoire of CD25+CD4 T cells is highly self-reactive, a conclusion supported by the observation that when T cells transduced with TCR genes derived from CD25+, but CD25− CD4 T cells, they rapidly expand in lymphopenic hosts and induce autoimmune disease (16). Furthermore, studies have shown that TCR transgenic T cells can undergo conversion into Treg cells following exposure to either cognate Ag or peptide-agonist ligands on dendritic cells, respectively (17, 18, 19).
Although a vast amount of literature exists on the developmental biology, function, and TCR repertoires of CD4+ Treg cells, details about CD8 Treg cells are only beginning to emerge. Studies on CD122-deficient mice have implicated the existence of CD8 Treg cells that function to preferentially regulate the immune functions of CD8 T cells. CD122-deficient mice exhibit severe hyperimmunity (9), which is associated with the expansion of abnormally activated T cells (10). However, the transfer of highly purified CD8+CD122+ T cells, from wild-type mice to CD122-deficient neonates, prevented the aberrant T cell phenotype from developing in the treated mice (11). Moreover, RAG-2−/− mice that received wild-type CD8+CD122− cells die within 10 wk after cell transfer, suggesting CD8+CD122− T cells become dangerously activated in the absence of CD8+CD122+ T cells (11). Follow-up studies indicate the suppressor activity of CD8+CD122+ T cells was mediated by IL-10 (12). Collectively, these results suggest CD8+CD122+ T cells contain novel populations that can function as Treg cells.
RasGRP1 is one of two Ras-guanyl-nucleotide exchange factors that link TCR signal transduction to Ras and MAPK activation (20, 21). Upon TCR stimulation, RasGRP1 mobilizes to the Golgi membrane by binding the phospholipase C-γ1 product diacylglycerol (DAG) through its C1 domain (22, 23, 24). Thymocytes from RasGRP1−/− mice are defective in TCR- and DAG-induced activation of Ras-ERK signaling (20). Furthermore, mutant mice exhibit a defect in positive selection, as evidenced by reduced numbers of single-positive (SP) thymocytes and T cell lymphopenia (25). By contrast, strong TCR signals responsible for negative selection and the induction of Ag-driven growth appear to be RasGRP1 independent (21). Because most Treg cells characterized to date express high affinity TCRs for self Ags, our objectives with regard to the analysis of the role of RasGRP1 in the development and function of Treg cells are 2-fold, as follows: 1) determine the role of RasGRP1 in the intrathymic development of CD4 and CD8 Treg cells, and 2) determine the role of RasGRP1 in peripheral homeostasis and function of CD4 and CD8 Treg cells. Our results indicate that although intrathymic development of CD4+Foxp3+ cells is severely impaired in the absence of RasGRP1, there exist an elevated frequency and large numbers of Foxp3-expressing CD4 Treg cells in the peripheral lymphoid tissues of mutant mice. This may be attributable to both the massive expansion of RasGRP1−/− Foxp3+CD4+ T cells and increased death rate of mutant Foxp3−CD4+ T cells. Additionally, RasGRP1−/− CD4 Treg cells were found to be functional because they could suppress the proliferation of wild-type CD25−CD4+ T cells in vitro. In contrast to the CD4 Treg cells, the development of CD8 Treg cells is not affected by RasGRP1 loss. However, the suppressor function of CD8 Treg cells is dependent on RasGRP1. The implications of these findings on peripheral T cell homeostasis and the development of autoimmune diseases in RasGRP1−/− mice are discussed.
Materials and Methods
C57BL/6J (B6) mice were obtained from The Jackson Laboratory. RasGRP1−/− breeder mice (20) were provided by J. Stone (University of Alberta, Edmonton, Alberta, Canada) and bred onto B6 background for more than seven generations. Mice 6–12 wk of age were used for the experiments described. All studies followed guidelines set by the Animal Care Committee at the University of British Columbia in conjunction with the Canadian Council on Animal Care.
Abs against CD4 (GK1.5), CD8α (53-6.7), CD8β (53.58), TCRβ (H57-597), CD3ε (2C11), CD25 (PC61.5), CD44 (IM7), CD62L (MEL-14), CD69 (H1.2F3), CD94 (18d3), CD122 (5H4), CD127 (A7R34), Foxp3 (FJK-16s), NK1.1 (PK136), NKG2AB6 (16a11), NKG2D (CX5), Thy1.1 (HIS51), Thy1.2 (53-2.1), and IL-10 (JES5-16E3) were purchased from eBioscience. Annexin V-PE and Abs against 2B4, CD5 (53-7.3), Ly6C (AL-21), and Ki-67 (B56) were purchased from BD Biosciences. Foxp3 staining was performed following the protocol recommended on eBioscience web site (www.ebioscience.com/ebioscience/specs/antibody_12/12-5773.htm). For anti-IL-10 staining, cells were stained for surface markers, washed, and fixed with 2% paraformaldehyde and 0.2% Tween 20 in PBS for 20 min on ice, followed by washing with PBS. Fixed cells were then stained with anti-IL-10 Ab in 0.2% Tween 20/PBS for 30 min on ice. For anti-Ki-67 staining (26), cells were fixed and permeabilized using the same protocol as for Foxp3 staining, and incubated with anti-Ki-67 Ab for 30 min at 4°C. Data were acquired using either FACScan/CellQuest software or LSRII/FACSDiva software (BD Biosciences). Data were analyzed with FlowJo (Tree Star) software.
CD4 T cell proliferation and suppression assays
Cell sorting with the FACSAria flow cytometer (BD Biosciences) was used to purify CD4+CD25− or CD4+CD25+ cells. The purities of sorted CD4+CD25+ wild-type or mutant suppressor cells used for experimental studies are 94.2 and 95.4%, respectively. Proliferation and suppression assays were performed, as described (27). Briefly, for proliferation assays, T cells (2 × 104 cells/well) were stimulated for 72 h with titrated amounts of Con A in the presence of T cell-depleted, irradiated APCs (8 × 104 per well) in 96-well round-bottom plates, and pulsed with 1 μCi per well of [3H]thymidine for the final 8 h. Suppression assays were performed under the same conditions using 2 × 104 CD4+CD25− T cells as responders, 8 × 104 irradiated APCs, and a 1:2 titration of the indicated suppressor T cell population (CD4+CD25+) at a starting concentration of 4 × 104 cells/well in the presence of Con A at 2 μg/ml−1 final concentration. All data are shown as mean [3H]thymidine incorporation in triplicate cultures.
Adoptive transfer experiment
Cell suspension was prepared from Thy1.1+ B6 or Thy1.2+ RasGRP1−/− animals. Cells were then stained with PE-conjugated anti-CD4 (GK1.5) and allophycocyanin-conjugated anti-CD25 (PC61.5) Abs and sorted for the CD4+CD25+ or CD4+CD25− population (purity >95%), respectively, using FACSAria. Sorted CD4+CD25+ (2 × 105) or CD4+CD25− (1 × 106) Thy1.1+ wild-type cells were adoptively transferred into naive B6 Thy1.2+ wild-type or RasGRP1−/− hosts. Similarly, sorted CD4+CD25+ (2 × 105) or CD4+CD25− (1 × 106) Thy1.2+ RasGRP1−/− cells were adoptively transferred into naive Thy1.1+ wild-type or RasGRP1−/− hosts. Spleens of recipients were recovered 3 wk posttransfer, and frequencies of donor cells of the indicated cell surface phenotype were quantified by flow cytometry.
Purified CD8+ T cells (1 × 107/ml) were labeled with 1 μM CFSE (Molecular Probes) in PBS for 10 min at room temperature. After stopping the reaction with the addition of an equal volume of FCS, cells were washed four times with complete medium.
CD8+ T cell purification and direct ex vivo assays
Single-cell suspensions from lymph nodes and spleens of mice were prepared and then treated with biotinylated anti-CD8 (53-6.7) mAb, followed by positive selection using MiniMACS system (Miltenyi Biotec), according to the manufacturer’s specifications. The resulting cells were >95% pure CD8+ TCR+ T cells. For cytokine proliferation assay, purified wild-type or RasGRP1−/− CD8+ cells were CFSE labeled and cultured in IL-2 (200 U/ml) or IL-15 (100 ng/ml). Proliferation of gated CD8+ cells was then analyzed by FACS at 72 h. For IFN-γ production assay, 2 × 106 purified wild-type or RasGRP1−/− CD8+ cells were stimulated with PMA (10 ng/ml) and ionomycin (100 ng/ml) in medium containing Golgi-plug (BD Pharmingen) for 5 h at 37°C. Following stimulation, cells were stained for surface Ags and then stained intracellularly for IFN-γ.
CD8+ T cell cytokine and suppression assays
CD8+ T cells from B6 wild-type and RasGRP1−/− mice (Thy1.2+) were purified, as described above. Cells were then stained with PE-conjugated anti-CD122 (5H4) and PE-Cy5-conjugated anti-CD8 (53-6.7) Abs and electronically sorted using FACSAria flow cytometer. Sorted wild-type CD8+CD122−, CD8+CD122+, or RasGRP1−/− CD8+CD122+ cells (5 × 104 per well) were stimulated by plate-bound anti-CD3 (10 μg/ml, 1-h incubation at 37°C, washed twice) and 10 μg/ml soluble anti-CD28 plus 100 U/ml IL-2 (final concentration) for 72 h in 24-well flat-bottom plates. For assessment of IL-10 production by the cultured CD8+ cells, Golgi-plug was added to the cell cultures during the last 24 h of culture and IL-10 production was evaluated by intracellular staining. For suppression assays, sorted wild-type CD8+CD122−, CD8+CD122+, or RasGRP1−/− CD8+CD122+ cells (2 × 106) were stimulated using the same condition above. Three days later, activated suppressor cells from the indicated population were added at various suppressor to responder ratios to 6.5 × 105 CFSE-labeled sorted B6 Thy1.1 CD8+CD122− cells, and cultured with anti-CD3 and anti-CD28 plus IL-2 (same stimulation condition as above) in 96-well flat-bottom plate. Two days later, the proliferation of Thy1.1+ CD8+CD122− wild-type cells was assessed by CFSE dilution analysis.
Defective thymic differentiation of CD4+Foxp3+ T cells in RasGRP1−/− mice
Because Foxp3 is a unique marker for naturally arising CD4+CD25+ Treg cells, we first examined the development of Foxp3+ CD4 SP thymocytes in the thymus of RasGRP1−/− mice. We found that although the proportion of CD4 SP thymocytes in mutant mice is 14-fold lower than in wild-type mice (0.44 vs 6.29%), the proportion of CD4 SP thymocytes that are also CD25+Foxp3+ in mutant mice is fairly similar to wild-type (2.02 vs 3.08%) (Fig. 1,A). However, cell number comparison of Foxp3+ or Foxp3− lineages in CD4 SP thymocytes between wild-type and RasGRP1−/− mice revealed that both CD4 Treg (Foxp3+) and non-Treg (Foxp3−) populations are severely affected by RasGRP1 deficiency, with both populations accounting for 4.5 and 4.2% of wild-type numbers, respectively (Fig. 1,B). Next, we examined the differentiation of Foxp3-expressing precursors in RasGRP1−/− thymus by evaluating the distribution of Foxp3-expressing thymocytes among the thymocyte subpopulations as defined by CD4 and CD8 expression. A recent report suggests that although Foxp3 induction can occur at the double-positive stage, it is preferentially induced at the CD4 SP stage during the development (28). We found that there is a significant increase in the proportion of CD4+CD8+Foxp3+ thymocytes in RasGRP1−/− mice relative to wild type (51 vs 13%) (Fig. 1,C). This result is consistent with the notion that there is a block in transition from Foxp3+ double-positive to Foxp3+ CD4 SP thymocytes in RasGRP1−/− mice. Alternatively, this observation could be due to a defect in Foxp3 up-regulation by RasGRP1−/− CD4 SP thymocytes. Interestingly, RasGRP1−/− Foxp3-expressing cells within the CD4 SP population display altered expression of TCRβ, CD3ε, CD5, and CD69 (Fig. 1,D). The distribution of TCR and CD5 in the Foxp3+ CD4 SP population is bimodal, with a minor population that expresses fairly normal levels of TCR and CD5 and a major population that expresses very low levels of these molecules (Fig. 1 D). If only the TCR+ population is representative of CD4 Treg cells, this would imply that the development of CD4 Treg cells is more greatly affected by the RasGRP1−/− mutation than implied by the analysis of Foxp3+CD4 SP thymocytes. In summary, these observations indicate that naturally arising CD4+Foxp3+ Treg cell development in the thymus is severely impaired in RasGRP1−/− mice.
Favored expansion and enhanced suppressive function of peripheral CD4+Foxp3+ T cells in RasGRP1−/− mice
In contrast to the impaired development of CD4+CD25+Foxp3+ Treg cells in the thymus, we found that the spleen of RasGRP1−/− mice possessed a markedly increased proportion of CD25+Foxp3+ cells within the CD4 population as compared with wild type (24 vs 8.6%) (Fig. 2,A). Furthermore, there is a preferential increase in splenic CD4+Foxp3+ cell number relative to the CD4+Foxp3− population in RasGRP1−/− mice (32 vs 14% of wild-type numbers; Fig. 2,B). We noted in Fig. 1,D that the majority of CD4+Foxp3+ thymocytes express very low levels of TCR and CD5. In striking contrast to the thymus, RasGRP1−/− peripheral CD4 Treg cells express near wild-type levels of TCRβ, CD3ε, and CD5 (Fig. 2,C). Interestingly, both Foxp3+ and Foxp3− CD4 T cells in RasGRP1−/− mice display signs of acute activation (CD44high, CD62Llow, CD69high), with a higher proportion of Foxp3+ CD4 cells from RasGRP1−/− mice expressing increased levels of acute activation markers (Fig. 2 C).
The huge increase in peripheral Foxp3+CD4 T cell numbers in RasGRP1−/− mice could be due to the preferential expansion of these cells and/or the preferential death of Foxp3−CD4 T cells in mutant mice. To test this hypothesis, we performed direct ex vivo staining of Foxp3+ and Foxp3− CD4 T cells to determine the proportion of proliferating and dying cells in the wild-type and mutant animals (Fig. 3,A). We found that both CD4 Foxp3− and Foxp3+ cells in RasGRP1−/− mice exhibit an elevated frequency of cells bearing the proliferation-associated nuclear Ag Ki-67 (26) as compared with wild-type animals (Foxp3−, 24.5 vs 7.2%; Foxp3+, 31.1 vs 15.6%) (Fig. 3,A, middle panel). Costaining of the apoptotic marker annexin V and Foxp3 revealed that Foxp3+ and Foxp3− CD4 T cells in RasGRP1−/− mice undergo different rates of cell death. It is noted that mutant Foxp3−CD4 T cells undergo a higher degree of cell death relative to their wild-type counterpart (13% annexin V+ vs 4%; Fig. 3,A). By contrast, there is a dramatic decrease in the proportion of mutant CD4+Foxp3+ cells that are annexin V+, as compared with wild-type mice (7.4% annexin V+ vs 22%; Fig. 3,A). To provide an explanation for the increased numbers/frequency of Foxp3+ T cells in the mutant periphery relative to the thymus, we performed a series of adoptive transfer experiments to distinguish between the cell intrinsic and extrinsic effects of the RasGRP1−/− mutation on the peripheral expansion of Foxp3+CD4 T cells (Fig. 3,B). First, sorted wild-type CD4+CD25+ or CD4+CD25− T cells (Thy1.1+) were transferred into either wild-type or mutant Thy1.2 hosts, respectively (Fig. 3,B, top panel). Three weeks postadoptive transfer, significant frequencies of donor CD4 T cells were detected only in RasGRP1−/−, but not wild-type recipients. This result suggests that it is the peripheral environment in RasGRP1−/− mice that favors the expansion of both wild-type CD4+CD25+ and CD4+CD25− cells. This is most likely a consequence of lymphopenia associated with RasGRP1−/− mice. We also determined whether there is increase conversion of donor CD4+CD25− into CD4+CD25+ T cells in RasGRP1−/− mice. We found that there is no conversion of either wild-type or mutant CD4+CD25− into CD4+CD25+ T cells in wild-type hosts (Fig. 3,C). Interestingly, 6.8% of donor wild-type CD4+CD25− cells developed into CD4+CD25+ cells in RasGRP1−/− host, and these CD4+CD25+ cells were also Foxp3+ (Fig. 3,C). Similarly, 4.9% of donor mutant CD4+CD25− T cells developed into CD4+CD25+Foxp3+ cells in mutant hosts (Fig. 3 C). Collectively, these findings suggest that the elevated frequency of CD4+CD25+Foxp3+ T cells in RasGRP1−/− mice is most likely due to a combination of decreased cell death and a cellular environment that favors their peripheral expansion.
To evaluate the function of CD4+Foxp3+ cells in the periphery of RasGRP1−/− mice, we purified both wild-type and mutant peripheral CD4+CD25+ T cells using FACS and compared their ability to suppress the proliferation of Con A-activated wild-type CD4+CD25− cells in vitro (Fig. 4). On a per cell basis, CD4+CD25+ T cells from RasGRP1−/− mice were more suppressive compared with wild-type CD4+CD25+ cells. This increased suppressor activity of CD4+CD25+ T cells from RasGRP1−/− mice correlates with the increased expression of CD44 and CD69 on these cells compared with their wild-type counterpart (Fig. 2 C). In summary, these studies demonstrate that there is preferential accumulation of functionally active CD4+Foxp3+ Treg cells in the periphery of RasGRP1−/− mice. They also indicate that the suppressor function of CD4+Foxp3+ Treg cells is independent of RasGRP1.
Preferential development of memory phenotype CD8+ T cells in the thymus and periphery of RasGRP1−/− mice
The Tec family tyrosine kinases Itk and Rlk are required for full TCR-induced activation of phospholipase C-γ1, Ca2+ mobilization, and ERK activation (29, 30). Itk and Rlk perform important functions during T cell development, and, in particular, have been implicated in setting the thresholds of positive and negative selection (31, 32, 33). Interestingly, although Itk and Rlk are critical for the development of conventional CD8 T cells, Itk−/− and Rlk−/−Itk−/− mice possess a large population of memory phenotype CD8 T cells that bear striking similarity to lineages of innate-like lymphocytes (34, 35, 36). Because RasGRP1−/− mice are defective in TCR- and DAG-induced activation of Ras-ERK signaling (20) and play a vital role in the positive selection of conventional T cells (20, 21), we question whether RasGRP1 is also dispensable for the development of these innate-like CD8 T cells that bear a memory phenotype. We found that there is a large increase in the proportion of CD44highCD122+CD8 SP thymocytes in RasGRP1−/− mice relative to wild type (Fig. 5,A, top panel). In addition, these CD122+ cells in the RasGRP1−/− thymus are CD24low, suggesting their mature status (Fig. 5,A, middle panel). There is also a slight increase in the total number of CD44highCD122+ cells in the thymus of RasGRP1−/− mice (115% of wild type) (Fig. 5,D). All of these observations contrast with the greatly reduced numbers and immature phenotype of CD4+Foxp3+ cells found in the thymus of RasGRP1−/− mice (Fig. 1,B). Consistent with the thymic data for this CD8 subset, there is ∼5-fold increase in the proportion of CD8+CD44highCD122+ cells in the spleen of RasGRP1−/− mice relative to wild type (Fig. 5,A, bottom panel). We also found that the proportions of CD8 SP thymocytes and splenocytes that are NK1.1+ are greatly increased in the thymus and spleen of mutant mice relative to the wild-type counterpart (Fig. 5,B). In absolute numbers, there is a 1.7- and 9-fold increase in the number of CD8+CD44highCD122+ and CD8+NK1.1+ cells, respectively, in the spleen of RasGRP1−/− mice as compared with wild type (Fig. 5,D). This observation contrasts with the significantly lower numbers of CD44lowCD8+ cells that are recovered from both the thymus and spleen of RasGRP1−/− mice relative to wild-type mice (Fig. 5,D). However, unlike CD4 Treg cells, there is no evidence that the elevation in the numbers of CD8+CD44highCD122+ T cells in RasGRP1−/− mice is due to either enhanced proliferation and/or decreased cell death of these cells relative to wild-type mice (Fig. 5 C). These observations suggest that CD8+CD44highCD122+ cells are more resistant to the effects of RasGRP1 deficiency than CD8+CD44lowCD122− T cells. Collectively, they indicate that there is a preferential development of memory phenotype CD8+CD44highCD122+ T cells in the thymus and spleen of RasGRP1−/− mice.
Elevated expression of NK receptors by CD8+ T cells in RasGRP1−/− mice
Functional differences between lymphocyte populations are often accompanied by changes in receptor expression patterns. It has been reported that CD8+CD44high cells from normal mice expressed significant levels of NK receptor upon IL-2 activation (37). In addition, CD44highCD122+ and CD44lowCD122− CD8+ T cells from Itk−/− and IL-15−/− mice, respectively, have distinct patterns of NK receptor expression (34). To distinguish the CD44highCD122+CD8+ T cells found in RasGRP1−/− mice from conventional memory CD8 T cells, we compared the expression of NK receptors and memory markers on CD8+ cells from wild-type and RasGRP1−/− mice. We found that the proportion of CD8+CD122+ cells expressing high levels of Ly6C, CD94, NKG2A/C/E, NKG2D, 2B4, and NK1.1 is dramatically increased in RasGRP1−/− mice compared with their wild-type counterparts (Fig. 6). This finding is similar to that previously reported for CD44highCD122+CD8+ T cells in Itk−/− mice (34). Notably, despite this dramatic difference in the expression of NK receptors, there are only minor differences in the expression of memory makers such as CD62L and CD127 between wild-type and RasGRP1−/− CD8+CD122+ spleen cells (Fig. 6). Therefore, it is likely that the CD44highCD122+CD8+ T cells present in RasGRP1−/− mice are similar to those described for Itk−/− mice and represent a cell lineage(s) that is distinct from conventional memory CD8 T cells.
Activated RasGRP1−/−CD122+CD8+ T cells are less suppressive than their wild-type counterpart
Recent studies (35, 36) have shown that CD44highCD122+CD8+ cells in Itk−/− mice can proliferate in IL-2 or IL-15 without TCR stimulation and produce IFN-γ directly ex vivo. To further characterize functional similarities of CD44highCD122+CD8+ T cells from RasGRP1−/− and Itk−/− mice, we determined whether these cells from RasGRP1−/− mice can proliferate in response to IL-15 or IL-2. As expected, naive conventional CD8+ T cells expressing low levels of CD122 did not proliferate when cultured with either IL-15 or IL-2. By contrast, CD8+CD122+ T cells from either wild-type or RasGRP1−/− mice proliferated vigorously when cultured with either IL-15 or IL-2 (Fig. 7,A). However, mutant CD8+CD122+ T cells proliferate less well than wild type in response to these cytokines, suggesting a role for RasGRP1 in the transmission of cytokine-dependent growth signals. Furthermore, only CD8+CD44high, but not CD8+CD44low T cells from both thymus and spleen of wild-type or RasGRP1−/− mice can produce IFN-γ after PMA plus ionomycin stimulation ex vivo (Fig. 7 B). These data indicate that the development of CD44highCD122+CD8+ T cells that possess innate immune functions is independent of RasGRP1.
Recent studies show that CD8+CD122+ cells from normal mice can also function as Treg cells via an IL-10-dependent mechanism (11, 12). We sought to investigate whether CD44highCD122+CD8+ T cells present in RasGRP1−/− mice possess similar suppressor function. To test this hypothesis, we compared the ability of CD8+CD122+ cells from RasGRP1−/− mice to produce IL-10 using wild-type CD8+CD122− cells as a negative control. Sorted CD8+CD122+ cells from wild-type or RasGRP1−/− mice were stimulated with anti-CD3 plus anti-CD28 and exogenous IL-2 for 3 days. IL-10 production was then evaluated by intracellular staining with anti-IL-10 mAb. As expected, wild-type CD8+CD122− cells failed to produce IL-10 after in vitro activation. By contrast, CD8+CD122+ cells from either wild-type or RasGRP1−/− mice produced significant amounts of IL-10 compared with wild-type CD8+CD122− cells (Fig. 8,A). We then compared the ability of activated CD8+CD122+ cells from wild-type or RasGRP1−/− mice to suppress the proliferation of wild-type CD8+CD122− cells (Fig. 8 B). Sorted CD8+CD122+ cells from wild-type or RasGRP1−/− mice were stimulated with anti-CD3 plus anti-CD28 and exogenous IL-2 for 3 days and used as a source of suppressor cells. Supernatants from these cultures were also collected and assessed for their suppressor activity at 1/2 dilution. CD8+CD122− cells and culture supernatants from wild-type mice activated in a similar manner were used as negative controls. CFSE-labeled wild-type CD8+CD122− (Thy1.1+) cells were used as responder cells. The Thy1.2+ activated suppressor (CD8+CD122+) cells and negative control (CD8+CD122−) cells were added at various suppressor to responder ratios and stimulated with plate-bound anti-CD3, anti-CD28 (10 μg/ml), plus 100 U/ml IL-2 for another 2 days. We found that activated CD8+CD122+ cells and culture supernatants from wild-type mice completely suppressed the proliferation of wild-type CD8+CD122− cells. By contrast, activated CD8+CD122+ cells from RasGRP1−/− mice were only suppressive when used at a high (1:5), but not lower ratios (1:15 and 1:45). Furthermore, the culture supernatants of activated mutant CD8+CD122− cells were not suppressive. These data suggest that although the development of memory phenotype CD8+CD44highCD122+ Treg cells is independent of RasGRP1, their suppressor function is much more dependent on RasGRP1.
We have previously concluded that RasGRP1 is especially crucial for transducing low-grade TCR signals because its absence preferentially affected the positive selection of the weakly selecting H-Y TCR (21). The development of CD4 T cells is also critically dependent on RasGRP1 (20, 25). By contrast, the development of thymocytes expressing the more strongly selecting 2C TCR and negative selection was much less sensitive to RasGRP1 loss (21). Because recent studies have suggested that a common denominator for the development of multiple Treg lineages is their dependence on high affinity TCR/self ligand interactions for their development (1, 8), we sought to determine how the development of Treg lineages was affected in RasGRP1−/− mice. In this study, we found that RasGRP1−/− mice can support the development of both CD4 and CD8 Treg cells, albeit with differences in efficacy in the generation of CD4 and CD8 Treg cells in the thymus of RasGRP1−/− mice.
In the absence of RasGRP1, development of CD4 Treg cells in the thymus is severely impaired, resulting in <5% yield relative to wild-type mice (Fig. 1,B). However, there is preferential expansion of CD4 Treg cells in the periphery, and the number of splenic CD4 Treg cells is ∼32% of wild type (Fig. 2,B). The development of CD4 Treg cells is influenced by at least two factors: availability of IL-2 and self Ag/MHC ligands (1, 38, 39). It remains to be determined whether the poor development of CD4 Treg cells in RasGRP1−/− thymus reflects the paucity of one or both of these factors. We have previously shown that T cells expressing relatively high affinity TCRs for self ligands can develop via a RasGRP1-independent mechanism (21). Consistent with this hypothesis is our observation that a large proportion of peripheral CD4 T cells in RasGRP1−/− mice is actively cycling and undergoing apoptosis, presumably as a result of high affinity TCR/self ligand interactions (40). In this study, we further characterized the CD4 phenotype in mutant mice by discriminating between Treg and non-Treg populations. Our observations suggest that in the periphery of RasGRP1−/− mice, although both Treg and non-Treg subsets are undergoing massive proliferation, a smaller proportion of Treg population shows signs of apoptosis as compared with wild type (Fig. 3,A), which might contribute to the high frequencies of CD4+Foxp3+ T cells in the mutant spleens. It is also conceivable that the proliferating CD4 non-Treg cells can produce cytokines that include IL-2. Together with the lymphopenia present in RasGRP1−/− mice, this could provide a favorable environment for peripheral CD4 Treg cell expansion that is mediated by its high affinity IL-2R. This hypothesis is supported by our adoptive transfer experiments demonstrating that there is a preferential expansion of adoptively transferred wild-type CD4+CD25+ cells in RasGRP1−/− mice relative to wild-type recipients (Fig. 3,B, top panel). Recent studies suggest that in the presence of cognate Ag, peripheral conventional CD4+CD25− T cells expressing high affinity TCRs for self ligands can be efficiently converted into Treg cells expressing Foxp3 (19). We therefore evaluated the possibility that mutant mice provide an environment niche that favors the conversion of conventional CD4+CD25− T cells to CD4+CD25+Foxp3+ Treg cells by transferring wild-type or mutant CD4+CD25− cells into wild-type or mutant recipients, respectively (Fig. 3,B). Interestingly, we found that the preferential expansion of transferred CD4+CD25− cells (either wild type or mutant) in RasGRP1−/− hosts is associated with an increase in the numbers of CD4+CD25+Foxp3+ cells (Fig. 3, B and C). This result raises the possibility that the increase in Foxp3-expressing CD4 T cells in the periphery of RasGRP1−/− mice could at least be due in part to the conversion of CD4+CD25− cells into CD4+CD25+Foxp3+ cells. However, because a small percentage of donor CD4+CD25− cells was Foxp3+ before transfer (see Fig. 2,B), the increase in the number of CD4+CD25+Foxp3+ cells in RasGRP1−/− hosts could either be due to the conversion of CD4+CD25−Foxp3+ cells into CD4+CD25+Foxp3+ cells or the conversion of CD4+CD25−Foxp3− cells into CD4+CD25+Foxp3+ cells. However, the fact that neither transferred mutant CD4+CD25+ or CD4+CD25− cells could survive in wild-type hosts (Fig. 3 B) indicates it is the mutant environment rather than cell intrinsic effects of the RasGRP1−/− mutation that plays a dominant role in the preferential increase of peripheral CD4 Treg cells in mutant mice.
In contrast to the development of CD4 Treg cells, intrathymic development of memory phenotype CD8 Treg cells is not affected in RasGRP1−/− mice. It is likely that the memory phenotype CD8 T cells that developed in RasGRP1−/− mice represent multiple lineages of CD8 T cells. Consistent with this hypothesis is the observation that these memory phenotype CD8 T cells are heterogeneous in terms of expression of cell surface markers such as CD44, CD122, NK1.1, Ly6C, CD94, NKG2A/C/E, NKG2D, and 2B4 (Fig. 6). These cells are also heterogeneous in IFN-γ production, with only 8.4 and 13.3% of total CD8 cells in the thymus and spleen, respectively, producing this cytokine upon PMA and ionomycin stimulation (Fig. 7,B), whereas 15 and 50.7% of total CD8 cells (thymus and spleen, respectively) possess the memory phenotype (CD44highCD122+) (Fig. 5 A). The preferential development of CD8 T cells that possess innate-like properties has also been observed in Tec kinase-deficient mice (34, 35, 36). The observation that CD8 T cells of similar cell surface phenotype and function develop in both RasGRP1 and Tec-kinase-deficient mice also suggests that RasGRP1 and Tec kinases activate components of the same signaling pathway. It is also likely that a common denominator in the generation of these memory phenotype CD8 T cells is the expression of relatively high affinity TCRs for self ligands, and therefore, these cells can develop via a RasGPR1-independent mechanism. Consistent with a previous report (12), we found that activated wild-type CD8+CD122+ cells are able to produce IL-10 and efficiently suppressed the proliferation of anti-CD3-, anti-CD28-, and IL-2-activated wild-type CD8+CD122− cells. Furthermore, culture supernatants derived from wild-type CD8+CD122+ cells activated with anti-CD3, anti-CD28, and IL-2 were also highly suppressive. However, similarly activated mutant CD8+CD122+ cells were much less suppressive at low suppressor to responder ratio. In addition, supernatants of activated mutant CD8+CD122+ cells failed to show detectable suppressive activity. Because we have previously shown that RasGRP1 plays a fundamental role in the developmental programming of Ag-activated CD8 T cells by providing signals necessary for their survival (21), it is possible that the less efficient suppressive activity of activated mutant CD8+CD122+ T cells may be due to the poorer survival of these cells. This poorer survival may also affect their ability to sustain the production of suppressor cytokines such as IL-10. Collectively, these results indicate that although RasGRP1 is not required for the development of CD8+CD44highCD122+ cells, the suppressive function of these cells is critically dependent on RasGRP1.
Consistent with our observation that peripheral lymphoid tissues of RasGRP1−/− mice harbor large numbers of both CD4 and CD8 Treg cells, we did not observe any overt autoimmune disorders in these mice, despite our hypothesis that only T cells expressing TCRs with high affinity for self ligands can develop in these mice. This observation contrasts with observations in a mouse strain that harbors a spontaneous mutation in the RasGRP1 (RasGRP1lag) gene (41). In RasGRP1lag mice, there are massive lymphoproliferation and development of an autoimmune syndrome that share similarities with systemic lupus erythematosus (41). It is possible that the autoimmune phenotype that developed in RasGRP1lag mice might be due to contributions of the hybrid 129:B6 background in addition to disruption of RasGRP1 gene function (42). Consistent with this hypothesis, we did observe massive lymphoadenopathy and splenomegaly in RasGRP1−/− mice in early backcrosses of 129 mutant mice to B6 mice. However, upon more than seven generations of backcross to the B6 background, no overt autoimmune symptoms were observed in RasGRP1−/− mice.
Our studies demonstrate that although the development of CD4 Treg cells in the thymus of RasGRP1−/− mice is very inefficient, peripheral mechanisms exist to greatly expand the numbers of CD4 Treg cells. It is likely that these CD4 Treg cells may be involved in preventing the development of overt autoimmune disease in RasGRP1−/− mice. By contrast, there is very efficient development of a heterogeneous population of memory phenotype CD8 T cells in the thymus, resulting in higher frequencies of these cells in the periphery of RasGPR1−/− mice. These memory phenotype CD8 T cells also include cells that can perform innate immune functions. To explain the contrasting functions of these memory phenotype CD8+ T cells that exhibit both innate functions and suppressive functions, we propose differential conditions used to activate memory phenotype CD8 T cells from either wild-type or RasGRP−/− mice lead to distinct immunological functions. The development of memory phenotype CD8 T cells in RasGRP1−/− mice may be due to high affinity TCR interactions with self Ags, because we have previously demonstrated the high expression levels of the memory markers CD44 and CD122 found in naive mice are maintained when the self Ag is present (43, 44). These studies complement those conducted in Itk−/− and Itk−/−Rlk−/− mice and support the hypothesis that unconventional CD8 T cells differ from conventional CD8 T cells in the requirement for Itk, Rlk, and RasGRP1 for their development. More importantly, our studies indicate RasGRP1 signaling plays a discriminative role in the intrathymic development of CD4 and CD8 Treg cells, whereas it is not required for peripheral expansion of either lineage.
We thank Soo-Jeet Teh for technical assistance, Salim Dhanji for insightful discussions, the Wesbrook Animal Unit for animal husbandry, and the Life Sciences Centre Flow Cytometry Facility for cell sorting.
The authors have no financial conflict of interest.
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.
This work was supported by grants from the Canadian Institutes of Health Research (Grant MOP-77547) and the National Cancer Institute of Canada with funds from the Terry Fox Foundation (Grant no. 016342) to H.-S.T.
Abbreviations used in this paper: Treg, T regulatory; DAG, diacylglycerol; SP, single positive.