Store-operated Ca2+ entry (SOCE) is believed to be of pivotal importance in T cell physiology. To test this hypothesis, we generated mice constitutively lacking the SOCE-regulating Ca2+ sensor stromal interaction molecule 1 (STIM1). In vitro analyses showed that SOCE and Ag receptor complex-triggered Ca2+ flux into STIM1-deficient T cells is virtually abolished. In vivo, STIM1-deficient mice developed a lymphoproliferative disease despite normal thymic T cell maturation and normal frequencies of CD4+Foxp3+ regulatory T cells. Unexpectedly, STIM1-deficient bone marrow chimeric mice mounted humoral immune responses after vaccination and STIM1-deficient T cells were capable of inducing acute graft-versus-host disease following adoptive transfer into allogeneic hosts. These results demonstrate that STIM1-dependent SOCE is crucial for homeostatic T cell proliferation, but of much lesser importance for thymic T cell differentiation or T cell effector functions.
The participation of ionized calcium (Ca2+) as a second messenger is an essential component of many cellular signaling events (1). In nonexcitable cells, including lymphocytes, Ca2+ is released from the endoplasmic reticulum (ER)4 via inositol 1,4,5-triphosphate (IP3)-mediated receptor activation triggered by ligand-activated plasma membrane receptors. If the limited Ca2+ reservoir of the ER becomes exhausted, extracellular Ca2+ enters the cytoplasm through calcium release-activated calcium (CRAC) channels in the plasma membrane by a mechanism referred to as store-operated calcium entry (SOCE) (2, 3). In T cells, SOCE triggered on binding of Ag to the TCR initiates Ca2+-dependent activation of transcription factors like NFATs, a process thought to be of paramount importance during T cell activation and proliferation (4, 5).
STIM1 has been identified as the “missing link” that connects intracellular store depletion to the opening of CRAC channels in T cells. STIM1 resides in the membrane of the ER, where the calcium binding EF-hand motif in its N-terminal region senses intraluminal calcium (6, 7, 8). After store depletion, STIM1 redistributes into puncta near the plasma membrane and activates CRAC channels (6, 7, 9, 10). In human T cells, the four-transmembrane domain protein Orai1 (also called CRACM1) is the predominant SOC channel (11, 12, 13, 14). Although SOCE has long been recognized as a major component of TCR-induced signaling, the knowledge about its significance for T cell development and function is limited. Mutation of Orai1 abolishes SOCE and has been associated with reduced production of cytokines, including IL-2 and IL-4, and defective T cell-mediated immune responses in vivo (14, 15). In line with this, the calcium-calcineurin-NFAT signaling pathway has been reported to be crucial for TCR-dependent expression of cytokines and several hundred other genes (16, 17) as well as thymocyte development (18). Furthermore, calcium signals may be required to maintain a stable “immunological synapse” between T cells and APCs (19) where STIM1 and Orai1 have, indeed, been shown to accumulate (20). A recent study using mice with a T cell-specific deficiency in STIM proteins confirmed the essential role of STIM1 for SOCE and CRAC function in T cells and revealed that stromal interaction molecule 2 (STIM2) contributes to sustained Ca2+ influx into those cells (21). The combined deficiency in STIM1 and STIM2 resulted in a defect in T cell homeostasis, which was mainly attributed to reduced numbers and impaired cellular functions of CD4+Foxp3+ regulatory T (Treg) cells. The contribution of STIM molecules to T cell development was not assessed, because gene deletion occurred at a late stage of T cell development.
Materials and Methods
The generation of Stim1−/− mice has been described previously (22). Generation of bone marrow (BM) chimeras was as follows: 5- to 6-wk-old C57BL/6 female mice were lethally irradiated with a single dose of 10 Gy, and BM cells from 4-wk-old wild-type (wt) or Stim1−/− mice were injected i.v. into the irradiated mice (4 × 106 cells/mouse). Eleven weeks after transplantation, lymph node cells and thymocytes were isolated and STIM1 deficiency was confirmed by western blotting using anti-STIM1 mAbs from BD Transduction (GOK/Stim1, clone 44) and Abnova (clone 5A2, H00006786-M01). All recipient animals received acidified water containing 2 g/L neomycin sulfate for 6 wk after transplantation. Rag−/− mice were bred at the Department of Neurology (University of Würzburg, Würzburg, Germany). CD90.1-congenic C57BL/6 (B6) and wt B6 mice were bred at the animal facility of the Institute for Virology and Immunobiology (University of Würzburg). BALB/c.OlaHsd hosts for acute graft-versus-host disease (aGvHD) experiments were obtained from Harlan-Winkelmann and used for experiments at 8 wk of age. All experiments were performed according to the Bavarian state regulations for animal experimentation and approved by the Regierung von Unterfranken as the responsible authority.
Processing of lymphoid tissues
Single-cell suspensions from thymi, lymph nodes, and spleens were prepared using a 70-μm nylon cell strainer (BD Falcon). Spleen cell suspensions were further subjected to hypo-osmotic shock for RBC depletion.
The following mAbs were used: anti-CD25, anti-CD44, anti-Gr1 FITC, anti-CD11b, anti-CD62L, anti-CD4, anti-CD90.1, anti-IFN-γ, and anti-Ki-67 PE or biotin, anti-CD8 PE-Cy5, anti-CD3 PE-Cy7, anti-CD4 Alexa Fluor 647 (all BD Pharmingen), and anti-Foxp3 PE-Cy5 (eBioscience).
Stainings were performed with up to 1 × 106 cells in 50 μl of PBS/0.1% BSA/0.02% NaN3. FcγRII/III receptors were blocked by incubation with saturating amounts of cell culture supernatant of the clone 2.4G2. Fluorochrome-conjugated or biotinylated mAbs were added after blocking (15 min, 4°C). Bound biotinylated Abs were detected by incubation with either PE-Cy5 or allophycocyanin-conjugated streptavidin (BD Pharmingen). The cells were analyzed on a FACSCalibur flow cytometer using CellQuest software (all BD Biosciences). Dot plots and histograms are shown as log10 fluorescence intensities on a four-decade scale.
For intracellular staining of Ki-67 and Foxp3 expression, cells labeled with Abs at the cell surface were fixed for 30 min at room temperature with fixation buffer (eBioscience) before permeabilization (permeabilization buffer; eBioscience). The cells were blocked with rat serum before staining with anti-Foxp3 mAb and anti-Ki-67 mAb for 30 min at room temperature. To determine IFN-γ expression, splenocytes were first restimulated with PMA/ionomycin (5/500 ng/ml; Sigma-Aldrich) for 4 h in the presence of GolgiPlug (1/1000; BD Biosciences) during the last 2 h of culture. Intracellular staining with anti-IFN-γ-PE was performed without the addition of rat serum.
Ca2+ measurements in primary T cells
Lymph node cell suspensions from two to three wt, wt BM chimeric (BMc), Stim1−/−, and Stim1−/− BMc mice were incubated with 15 μl/5 × 106 cells of a mix of 25 μl of Indo 1 (2 mM in DMSO; Molecular Probes), 25 μl of pluronic (Molecular Probes), and 113 μl of FCS (Life Technologies) for 45 min at 37°C. Subsequently, cells were stained with anti-B220 PE and anti-CD8 PE. For TCR complex stimulation, aliquots of cells were loaded with anti-CD3 mAb (clone 145-2C11; BD Pharmingen) at 10 μg/ml for 15 min on ice with 2 × 107 cells/ml. After a washing step, cells were resuspended in cell culture medium (see below). Before the initiation of Ca2+ measurements, cells were diluted 1/2.5–1/5 in cell culture medium prewarmed to 37°C or pelleted cells were resuspended in prewarmed Ca2+-free Tyrode’s buffer and allowed to equilibrate for 5 min at 37° C. Baseline Ca2+ levels were recorded for 50 s before CaCl2 (1 or 3 mM) was added. Bound anti-CD3 mAb was cross-linked by addition of goat anti-hamster Ig FITC (Dianova) at 20 μg/ml. Ionomycin (Sigma-Aldrich) was used at a final concentration of 1 μg/ml and thapsigargin (TG; Molecular Probes) at 5 μM. Data were acquired on a FACSDiva using CellQuest (both BD Biosciences) and FlowJo (Tree Star) was used to calculate the kinetics of FL5/FL4 means (Gaussian smoothing) among CD8−B220− lymph node cells (>93% CD4+ cells).
Western blot analysis of STIM1 expression in lymph node cells
Whole lymph node cells were lysed using 1% Nonidet P-40. Lysates of 1 × 106 cells were used for SDS-PAGE protein separation followed by protein blotting onto a polyvinylidene difluoride membrane. The anti-STIM1 Ab was from Abnova (H00006786-M01).
Serum samples and culture supernatants (see below) were subjected to a cytometric bead array (CBA) analysis (BD Pharmingen) using a mouse Th1/Th2 cytokine kit according to the manufacturer’s instructions. Samples were diluted 1/4 and primary data were analyzed using the BD Pharmingen CBA software.
In vitro suppression assays
CD4+ T cells were purified from lymph nodes of wt or Stim1−/− BMc mice by negative magnetic separation (see below). CD25+ cells were isolated by staining with anti-CD25 FITC (BD Pharmingen) and anti-FITC beads (enrichment) and passage over an LS column (Miltenyi Biotec). The magnetic fraction was again passed over an LS column to increase purity of CD4+CD25+ cells. CD25− cells were obtained by passing the nonmagnetic fraction from the first LS column separation again over an LD column (Miltenyi Biotec). CD4+CD25+ were 81% Foxp3+ and CD4+CD25− were 92% Foxp3−. To measure suppression, either wt or Stim1−/− CD4+CD25− cells (2.5 × 104 cells/well) were cocultured with 2.5, 0.5, or 0.1 × 104 CD4+CD25+ cells (Treg cells) per well or the same numbers of CD4+CD25− indicator T cells (Tind) in the presence of 2.5 × 104 irradiated splenocytes (B6) and 0.5 μg/ml anti-CD3 mAb in a 96-well U-bottom plate with a final volume of 100 μl. All cultures were in triplicates. Proliferation was measured by determining [3H]thymidine incorporation (12.5 mCi of [3H]thymidine; GE Healthcare) during the final 16 h of a 3-day culturing period. Radioactive incorporation was determined by beta counting. Percent suppression was calculated as: 100 × (1 − (a + n)/a × mean cpm (a × Tind + n × Treg)/mean cpm ((a + n) × Tind)), where a is the number of Tind per well and n is the number of Treg cells per well.
In vitro thymocyte apoptosis assays
Thymocytes were isolated from wt or Stim1−/− BMc mice and seeded at 2 × 105 cells/well of a 96-well U-bottom plate (Greiner). Anti-CD3 mAb or dexamethasone (Sigma-Aldrich) was added at different concentrations. After overnight culture, cells were stained with anti-CD4 PE and anti-CD8 PE-Cy5. After a washing step, cells were preincubated for 15 min on ice with annexin V-FITC (BD Pharmingen), which was diluted 1/8 with PBS/0.1% BSA/0.02% NaN3 immediately before analysis.
T cell transfers into Rag1-deficient mice
CD4+ T cells were purified (see below) from RBC-lysed splenocytes and lymph nodes of either wt or Stim1−/− BMc mice and 5 × 106 purified CD4+ cells were transferred i.v. into Rag-deficient recipients.
Anti-CD3 stimulation in vivo
Fifty micrograms of anti-CD3 mAb (BioLegend) or PBS as control was injected i.p. into wt or Stim1−/− BMc mice. Mice were sacrificed after 3 days to perform FACS analyses.
Keyhole limpet hemocyanin (KLH) immunizations
Wild-type or Stim1−/− BMc mice were immunized s.c. with an emulsion of 45 μl of PBS containing 25 μg of KLH (Sigma Aldrich) and 15 μl of TiterMax (Alexis). Twenty days after the initial immunization, mice were boosted with 25 μg of KLH in 250 μl of PBS i.p. Blood was taken on days 16 and 34 and anti-KLH Abs were detected in serum by ELISA using KLH-coated plates (150 μg/ml in PBS at a final volume of 100 μl/well). Wells were blocked with BSA (5% in PBS) before serum was added. Bound anti-KLH Abs were detected with anti-mouse IgG-biotin (Dianova) followed by streptavidin-peroxidase (Dianova) and tetramethylbenzidine (The Binding Site) was used as substrate. Anti-KLH Abs were subspecified into different Ig isotype classes by detecting bound anti-KLH Abs with anti-mouse IgG2a-peroxidase, anti-mouse IgG1-peroxidase (Dianova), or anti-mouse IgE-biotin (BD Pharmingen). ODs were measured at 450 nm. On day 43 or 63, CD4+ T cells were purified by negative magnetic separation from total splenocyte suspensions and 5 × 104 or 1 × 105 CD4+ cells were cocultured with 5 × 104 or 1 × 105 irradiated splenocytes (B6) in 96-well U-bottom plates. Cultures were in replicates of six. KLH was added at a final concentration of 10 μg/ml. Supernatants for detection of cytokines were collected and frozen after 48 h. Cytokine secretion was quantitated by CBA according to the manufacturer’s instructions. Cytokine concentrations from supernatants of unstimulated cells were subtracted from those of KLH-stimulated CD4+ T cells and the result was divided by the mean cytokine concentration of unstimulated cells from both groups.
Serum Ig concentrations
Concentrations of Abs of the IgM, IgG2a, IgG1, and IgE isotype were determined by ELISA according to the manufacturer’s (Bethyl Laboratories) recommendations. ODs were translated into protein concentrations using GraphPad Prism 4.0c.
Induction of aGvHD
BALB/c mice were conditioned for transplantation by total body irradiation with 8 Gy as a single dose. To prevent bacterial infections, animals were given neomycin (250 μg/ml; Sigma-Aldrich) and polymyxin B sulfate (3 U/ml; Sigma-Aldrich) in drinking water starting 3 days before irradiation until day 28 after transplantation. Approximately 24 h after irradiation, the mice received 1 × 107 T cell-depleted (TCD) BM cells from B6 mice and 5 × 105 CD4+ T cells from wt or Stim1−/− BMc mice. To obtain TCD BM cells, erythrocytes were lysed from total BM preparations, then FcγRs were blocked with 20 μg/ml normal mouse Ig (Sigma-Aldrich) before T cells were depleted using MACS anti-CD90.1 biotin (BD Pharmingen) followed by streptavidin beads (Miltenyi Biotec) and MACS separation columns according to the manufacturer’s instructions. T cell depletion was ∼95% on average. CD4+ T cells were purified from RBC-lysed lymph node cells with average purities of 90% by negative magnetic selection of cells expressing CD8a, CD11b, B220, CD49b, and/or Ter-119 (Miltenyi Biotec). Observers blinded to the treatment measured body weight and scored clinical appearance of the animals every other day as follows (scores in parentheses) (23): weight loss: <10% (0), >10 to <25% (1), and >25% (2); posture: normal (0), hunching noted only at rest (1), and severe hunching impairs movement (2); activity: normal (0), mild to moderately decreased (1), and stationary unless stimulated (2); fur texture: normal (0), mild to moderate ruffling (1), and severe ruffling/poor grooming (2); and skin integrity: normal (0), scaling of paws/tail (1), and obvious areas of denuded skin (2). At each observation time point, values were added to obtain one cumulative clinical scoring value per animal. Mice with <70% of the initial body weight for more than 2 days were euthanized. Alternatively, animals were killed independently of their body weight to prevent severe suffering as indicated by their overall clinical appearance.
Values of p given either in the figures or in the text are the results of two-tailed Student’s t tests (Microsoft Excel 11.3.5) assuming equal variance within groups. Two-way ANOVA testing (GraphPad Prism 4.0c) was used as indicated. A value of p < 0.05 was considered statistically significant.
Development of mature T cells in STIM1-deficient mice
To investigate the role of STIM1 in T cell development and effector functions, we analyzed mice constitutively lacking the protein. As recently reported, STIM1 deficiency is associated with ∼70% perinatal lethality, probably related to a cardiopulmonary defect. The surviving animals displayed pronounced growth retardation and a maximal life span of 4–6 wk (22, 24). Thymi of Stim1−/− mice were macroscopically and histologically normal, although the absence of STIM1 protein was confirmed by Western blot analysis (Fig. 1,A and data not shown). Furthermore, although absolute numbers of thymocytes were about 2-fold reduced (data not shown), thymocyte subset composition as determined by flow cytometry was comparable between Stim1−/− mice and wt littermate controls (Fig. 1,B). Similarly, lymph nodes of Stim1−/− mice also showed normal frequencies of both CD4+ and CD8+ cells (Fig. 1 B).
To analyze the impact of STIM1 deficiency on SOCE, we stimulated primary T cells from Stim1−/−mice and wt littermates with the sarco-/ER Ca2+ ATPase inhibitor TG (Fig. 1,C). Although the TG-induced Ca2+ influx into wt CD4+ T cells (Fig. 1,Cc) was almost as pronounced as that observed with the Ca2+ ionophore ionomycin (Fig. 1 Ca), CD4+ T cells from Stim1−/− mice showed a greatly reduced Ca2+ response to ionomycin (Fig. 1,Cb) and no detectable response to TG (Fig. 1,Cd). When the T cells were kept in Ca2+-free buffer, Ca2+ release from internal stores was clearly measurable after TG stimulation in T cells from both wt (Fig. 1,C, e and g) and Stim1−/− mice (Fig. 1,C, f and h). The consecutive SOCE after addition of 1 or 3 mM extracellular Ca2+, however, was almost completely abolished in Stim1−/− T cells as compared with wt controls. Similarly, Ca2+ flux into anti-CD3- stimulated T cells from Stim1−/− mice (Fig. 1,Cj) was much lower than that into wt T cells (Fig. 1 Ci) and barely above baseline. This result confirmed that SOCE is abolished in murine T cells in the absence of STIM1 and that this defect cannot be compensated by STIM2 which is also expressed in these cells (21).
Taken together, STIM1 deficiency abolishes TG-induced SOCE and Ag receptor complex-triggered Ca2+ fluxes into T cells, but it does not have a major effect on thymic T cell maturation and allows for normal frequencies of CD4+ and CD8+ T cells in lymph nodes.
Lymphoproliferation despite normal frequencies of functional Treg cells in Stim1−/− and Stim1−/− BMc mice
Upon necropsy, the most prominent characteristic of Stim1−/− mice was a pronounced splenomegaly (Fig. 1,D). Flow cytometric analyses revealed the accumulation of a population of FSClowSSChighCD11b+Gr1low myelomonocytic cells in the spleens of the mutant animals, which constituted up to 30% of all splenocytes (Fig. 1 E). Cytospins of FACS-sorted FSClowSSChigh cells from the mutant mice showed that the majority of these cells were eosinophils, while in control mice FSClowSSChigh splenocytes mainly consist of neutrophils (supplemental Fig. S1A).5 Histological assessment of nonlymphoid tissues further showed leukocytic infiltrates in the lungs and the livers, among other tissues, of Stim1−/− mice (supplemental Fig. S1B).
In contrast to lymph nodes (Fig. 1,B), splenic lymphocytes of Stim1−/− mice harbored a higher proportion of CD4+ T cells than wt controls (Fig. 1,B), which were also highly activated as determined by low CD62L and CD3 expression and high CD44 expression (Fig. 2, A and B). Although only ∼26% of splenic CD4+ T cells from wt mice displayed an effector/memory cell phenotype (CD62L−CD44+), ∼71% of CD4+ T cells from Stim1−/− mice were effector/memory cells (Fig. 2,A). In contrast, there was no difference in the percentages of effector/memory cells among CD4+ T cells from lymph nodes of wt controls and Stim1−/− mice. The high percentage of effector/memory-like cells was accompanied by a 2-fold lower expression of CD3 at the cell surface of Stim1−/− T cells as compared with wt controls (Fig. 2 B), which suggested that the hyperactivated phenotype of Stim1−/− T cells was a consequence of TCR stimulation in vivo (25).
Because CD25+/−Foxp3+ Treg cells constrain T cell activation under noninflammatory conditions in normal mice (26) and as ablation of STIM1 and STIM2 in double-positive thymocytes has been shown to severely hamper Treg cell generation (21), we determined the frequency of Foxp3+ cells among CD4+ cells of Stim1−/− mice. As shown in Fig. 2,C, Stim1−/− mice had normal frequencies of CD4+Foxp3+ cells, which also displayed an activated phenotype as indicated by high CD25 expression. Among CD4+Foxp3− cells, however, the differences in CD25 expression were even more pronounced as the percentage of CD25+ cells was about 10-fold higher in Stim1−/− mice than in wt controls. Analysis of the expression of the proliferation marker Ki-67 among CD4+ cells further revealed that both Foxp3− and Foxp3+CD4+ cells from Stim1−/− mice had a much higher turnover rate in vivo than cells from wt controls (Fig. 2,D). Constitutive T cell activation and proliferation in Stim1−/− mice were, however, not accompanied by increased cytokine levels in the circulation (Fig. 2 E).
For further studies on immune functions in the absence of STIM1 and SOCE, we generated BMc mice by reconstituting wt B6 mice with BM cells from either Stim1−/− or wt littermate control mice because Stim1−/− mice died between 4 and 6 wk of age (22). The lymphoproliferative phenotype of Stim1−/− mice was recapitulated in Stim1−/− BMc mice, again in the presence of normal Treg cell frequencies (Fig. 3, A and B). Therefore, we performed crisscross in vitro suppression assays testing for the capacity of lymph node CD4+CD25+ Treg cells from wt or Stim1−/− BMc animals to suppress either wt or Stim1−/− conventional T cells, i.e., we cocultured different numbers of purified CD4+CD25+ Treg cells with a fixed number of CD4+CD25− conventional T cells and stimulated the T cells with anti-CD3 in solution in the presence of APCs. As depicted in Fig. 3,C, Treg cells from Stim1−/− BMc mice were equally potent in suppressing both wt and Stim1−/− conventional T cells as wt Treg cells. Of note, the higher degree of suppression of Stim1−/− conventional T cells (Fig. 3,C, right graph) in comparison to wt T cells (Fig. 3,C, left graph) can be attributed to the 5-fold lower overall proliferation of Stim1−/− conventional CD4+CD25− T cells under these conditions in vitro (Fig. 3 D).
Apart from peripheral tolerance mechanisms, other mouse mutants with impaired TCR signaling displayed a lymphoproliferative or autoimmune phenotype due to altered negative selection in the thymus (27). We assessed the susceptibility of double-positive thymocytes from wt or Stim1−/− BMc mice to undergo negative selection in vivo by determining the degree of anti-CD3-induced apoptosis in vitro: Double-positive thymocytes from wt and Stim1−/− BMc mice were equally susceptible to anti-CD3-induced apoptosis (Fig. 4, right graph) and also showed a similar degree of dexamethasone-induced apoptosis (Fig. 4, left graph), which we used as a positive control to determine susceptibility of Stim1−/− thymocytes toward apoptotic stimuli in general.
Thus, Stim1−/− and Stim1−/− BMc mice developed a lymphoproliferative disease, although there was no obvious impairment of central tolerance mechanisms and normal frequencies of functional Treg cells in the periphery.
Increased homeostatic and conserved anti-CD3-induced proliferation of Stim1−/− T cells in vivo
The increased CD4+ T cell proliferation we had observed in Stim1−/− and Stim1−/− BMc mice was probably a consequence of increased proliferation to so-called homeostatic stimuli (28). To directly test for the reactivity of Stim1−/− T cells to homeostatic stimuli, that is autoantigen, Ag from commensal gut bacteria, and lymphopenia, we isolated CD4+ T cells from wt or Stim1−/− BMc mice and adoptively transferred them into Rag-deficient animals (29). Following T cell expansion in peripheral blood over a period of 7 wk (Fig. 5,A) and quantification of the progeny of transferred T cells in the spleens of recipient animals at the end of the experiment (Fig. 5,B) revealed an at least 3-fold higher expansion of CD4+ T cells from Stim1−/− mice as compared with wt controls (Fig. 5, A and B). Moreover, transfer of Stim1−/− CD4+ cells led to increased frequencies of host-derived CD11b+Gr1low cells in Rag-deficient mice (Fig. 5 C), indicating that the expansion of this cell population in Stim1−/− mice was, indeed, the result of the hyperactivation and hyperproliferation of CD4+ T cells in these animals.
During homeostatic proliferation, the lack of SOCE after CD3 stimulation in Stim1−/− T cells (Fig. 1,D and supplemental Fig. S2) could have been compensated for by an increased reactivity toward other stimuli of homeostatic proliferation. Therefore, we injected an anti-CD3 mAb into wt and Stim1−/− BMc mice to directly monitor proliferation to TCR complex stimulation in vivo (30) (Fig. 5,D). This treatment led to equal proportions of Ki-67+ cells among wt and Stim1−/− CD4+ or CD8+ T cells (Fig. 5 D), while overall T cell numbers were reduced to 60% (wt) and 70% (Stim1−/− BMc) of control-treated animals (data not shown). The increase in the percentages of Ki-67+ cells after anti-CD3 stimulation in vivo was greater for wt T cells than for Stim1−/− T cells, which mainly reflected the higher constitutive proliferation of Stim1−/− T cells as compared with wt T cells. Together, these results show that lack of SOCE in T cells leads to increased proliferation toward homeostatic stimuli and does not have a major effect on the proliferative response to anti-CD3 stimulation in vivo.
Stim1−/− T cells are capable of providing B cell help in vivo
We followed up on the conserved reactivity of T cells from Stim1−/− BMc mice to anti-CD3 treatment in vivo and analyzed the capacity of Stim1−/− T cells to provide help to STIM1-deficient B cells in vivo. B cells were present in normal frequencies in Stim1−/− animals, but also showed increased turnover as determined by Ki-67 expression (data not shown). Probably due to this constitutive B cell activation followed by continuous plasma cell differentiation, concentrations of Igs were up to 10-fold increased in sera of Stim1−/− BMc as compared with wt BMc mice with a bias toward TH2-dependent isotypes (supplemental Fig. S3A). However, auto, that is, anti-nuclear, Ab titers were only about 2-fold higher in the sera of Stim1−/− BMc mice as compared with wt BMc mice (supplemental Fig. S3B). Immunization of mice with the T cell-dependent model Ag KLH elicited a similar humoral immune response in wt and Stim1−/− BMc mice as indicated by slightly higher titers of KLH-specific Abs in the sera of Stim1−/− vs wt BMc mice after primary (Fig. 6,A, left graph) and equally high titers after booster immunizations (Fig. 6,A, right graph). Subspecification of anti-KLH Abs into different Ig isotypes revealed equal titers of the Th2-dependent isotype IgG1 in wt and Stim1−/− BMc mice, while Ag-specific IgE levels were slightly higher in sera of Stim1−/− BMc mice (Fig. 6,B). Stim1−/− BMc mice further showed a tendency toward lower titers of Th1-dependent Abs of the IgG2a isotype (Fig. 6,B). We sacrificed these animals after the boost to measure the CD4+ T cell response to KLH in in vitro recall assays to KLH. Splenic CD4+ T cells from Stim1−/− BMc mice secreted similar amounts of TNF-α and, in tendency, less IFN-γ into the supernatants as compared with wt CD4+ T cells, whereas we could not detect IL-4 production by CD4+ T cells from BMc mice of either genotype (Fig. 6 C). These results demonstrate that STIM1-dependent SOCE is dispensable for the differentiation of CD4+ T cells into helper T cells and the induction of a vigorous humoral immune response to protein Ag in vivo.
Stim1−/− CD4+ T cells induce graft-vs-host disease
Because during a humoral immune response in TCR- nontransgenic mice effector T cell function can only be measured by Ag recall in vitro (Fig. 6,C), we used a model of aGvHD to directly test for the capacity of Stim1−/− T cells to exert T cell effector functions in vivo. Thus, we transplanted TCD BM cells along with CD4+ T cells from wt or Stim1−/− BMc into fully allogeneic BALB/c mice. After T cell transfer recipients of both wt and Stim1−/− T cells developed severe aGvHD (Fig. 7,A). In recipients of Stim1−/− CD4+ T cells, disease onset (Fig. 7,A) and lethality (Fig. 7,B) were, however, delayed or reduced, respectively, indicating a somewhat lower capacity of these cells to induce T cell-mediated pathology in comparison to wt CD4+ T cells. Inbred B6 mice receiving CD4+ T cells from wt or Stim1−/− BMc mice did not develop aGvHD (Fig. 7, A and B). Therefore, the induction of aGvHD by Stim1−/− CD4+ T cells in BALB/c mice truly reflected the TCR-mediated response of Stim1−/− CD4+ T cells to alloantigen in vivo.
To follow up on the lower capacity of Stim1−/− vs wt CD4+ T cells to induce lethal aGvHD, we measured the concentrations of proinflammatory cytokines in the sera of recipient BALB/c mice 6 days after transplantation. Circulating levels of TNF-α and, in particular, IFN-γ, which both play central roles in the pathogenesis of aGvHD (31), were markedly diminished after transfer of Stim1−/− as compared with wt CD4+ T cells (Fig. 7,C). However, because Stim1−/− donor CD4+ T cells had also accumulated about 2-fold less than their wt counterparts (Fig. 7,D), the median cytokine concentration induced per cell was quite similar between groups for TNF-α, but not for IFN-γ (Fig. 7,E). Also, at the single cell level, as analyzed by flow cytometry, IFNγ production by Stim1−/− CD4+ T cells was low, but clearly detectable and, again, significantly less than for wt CD4+ T cells (Fig. 7 F).
In summary, Stim1−/− T cells, despite the absence of a measurable Ca2+ response after TCR complex stimulation in vitro, not only provided B cell help in vivo, but were also capable of inducing aGvHD upon transfer into allogeneic recipients.
The identification of STIM1 as the long-sought molecular link between Ca2+ store depletion and CRAC channel activation has enabled gain-of function (32) and loss-of-function (22, 24) studies to examine the significance of SOCE for mammalian physiology and disease processes. Although SOCE has been implicated in negative selection of thymic T cell precursors (33), Stim1−/− mice had a normal composition of thymocyte subsets. In the periphery, splenomegaly due to increased numbers of granulocytes and a relative increase in CD4+ T cells was the most obvious characteristic of Stim1−/− mice. Although splenic T cells contained very high frequencies of activated cells, lymph node T cells predominantly had a naive phenotype. This difference could, at least in part, be due to the inability of activated CD62Llow T cells to reenter into the lymph nodes via high endothelial venules. Alternatively, hyperactivation of eosinophils and secondary T cell activation might be the driving force behind the leukocytosis in Stim1−/− animals. Transfer of CD4+ T cells into Rag-deficient mice, however, revealed that these Stim1−/− CD4+ T cells not only produced higher amounts of IL-5 as compared with wt CD4+ T cells after anti-CD3 stimulation in vitro (data not shown), but also drew SSChighCD11b+Gr1low cells, i.e., putative eosinophils, into the spleen. It is possible that the myeloid population expanded in the spleens of Stim1−/− mice might, at least, in part overlap with the recently identified population of myeloid cells induced in the spleen after vaccination with alum as adjuvant (34), which could secondarily enhance T cell activation due to the primary defect in STIM1 expression and SOCE in T cells.
T cell activation is followed by production of cytokines of which IL-2 had originally been identified as the T cell growth factor in vitro. Similarly to the phenotype of Stim1−/− mice described here, it came as a surprise that IL-2-deficient mice (35) were not overtly immunodeficient, but developed a lymphoproliferative disease sharing many characteristics with that of Stim1−/− mutants (36). Although in IL-2-deficient mice lymphoproliferation evolves mainly due to the inability to maintain a functional Treg cell compartment exerting immunoregulation in trans, which is further aggravated by a reduced susceptibility of IL-2-deficient T cells to CD95-mediated cell death in cis (36), it seems that in Stim1−/− mice the primary deficiency is located in cis. Although we cannot exclude that Stim1−/− Treg cells might be slightly less efficient suppressors than wt Treg cells in vivo, Stim1−/− mice had normal frequencies of Treg cells and Treg cells from Stim1−/− BMc mice potently suppressed both wt and Stim1−/− T cells in vitro. This is in contrast to the recently published findings by Oh-hora et al. (21) in mice with a deletion of STIM1 and STIM2 in T cells, where the strongly diminished Treg cell compartment sufficed to explain the observed lymphoproliferation.
Because autoreactive T cells need costimulation through CD28 to produce IL-2 (37, 38), the degree of costimulation received under noninflammatory conditions is comparatively weak and autoreactive T cells might even recognize peptide-MHC complexes containing their “cognate” autoantigenic peptide without a costimulatory signal. Isolated stimulation of the TCR on purified T cells in vitro or culture of purified T cells in the presence of ionomycin, however, have been known for a long time to induce a state of anergy (39). The significance of the anergic phenotype observed in vitro for immune cell homeostasis in vivo has, however, remained elusive. Our data from STIM1-deficient mice suggest that there is, indeed, a SOCE-dependent negative feedback loop in T cells in vivo, which may probably not induce anergy as such, but which constrains the activation of autoreactive T cells. Moreover, the differential requirements of CD4+ and CD8+ T cells for anergy induction (40) might account for the stronger impact of STIM1 deficiency on CD4+ than on CD8+ T cells.
In contrast to autoreactive T cell activation in a noninfected host, immune responses to foreign Ag during bacterial or viral infections or after vaccination with Ag in adjuvants occur in the presence of ample costimulation. Although Stim1−/− T cells had a greatly reduced Ca2+ response after anti-CD3 stimulation in vitro, Stim1−/− and wt BMc mice mounted an equally vigorous Ab response to KLH in vivo. Production of IFN-γ and TNF-α by Stim1−/− T cells after restimulation with Ag in vitro and even more so, production of anti-KLH Abs of the IgG2a isotype revealed that Stim1−/− T cells had differentiated into Th1 cells. The degree of Th1 differentiation by STIM1-deficient T cells further suggested that Stim1−/− T cells had, if anything, received only a mildly “weaker” TCR signal as compared with wt T cells, as upon weak TCR stimulation one would expect Th2 polarization of the immune response by default (41). Due to the germline knockout mutation of the Stim1 gene, also B cells of STIM-deficient mice were defective in SOCE and anti-BCR-induced Ca2+ flux (data not shown). The Ab response of Stim1−/− BMc mice to KLH immunization in vivo, however, indicated that, like Stim1−/− T cells, Stim1−/− B cells were functional and capable of differentiating into Ab-secreting plasma cells and also memory B cells.
Although prophylactic vaccinations with bacterial or viral Ags against infectious diseases are very rarely accompanied by immunopathological complications, there is a high incidence of aGvHD after allogeneic BM transplantation. SOCE has also been implicated in the allostimulation of T cells in vivo (4), which induce aGvHD through the action of the effector molecules TNF-α and CD178 (42). Our data, however, suggest that STIM1-deficient T cells were almost as potent in inducing aGvHD as wt T cells. Therefore, activation of T cells in response to alloantigen in vivo and differentiation into effector T cells was not severely hampered by SOCE deficiency. This, in turn, indicates that inhibition of IP3 generation, for example, through interference with the function of coronin1 (43) upstream of SOCE or blockade of the Ca2+ calcineurin pathway by the most widely used immunosuppressant in aGvHD, cyclosporin A, downstream of SOCE might constitute more efficient means for the protection from aGvHD.
Taken together, our analyses on the functionality of the immune system of STIM1-deficient and STIM1-deficient BMc mice have revealed that not only T cell maturation, but also differentiation of T and B cells into effector cells can occur largely independent of SOCE.
We thank Nadine Pfeifer and Sandra Werner for conducting experiments and providing excellent technical assistance, Christian Linden for performing Ca2+ flux measurements, and Thomas Hünig for critical reading of this manuscript.
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 Rudolf Virchow Center, the Deutsche Forschungsgemeinschaft (Ni 556/7-1 to B.N.), and the Wilhelm Sander-Stiftung (2005.133.1).
Abbreviations used in this paper: ER, endoplasmic reticulum; aGvHD, acute graft-versus-host disease; BMc mice, bone marrow chimeric; CRAC, calcium release-activated calcium; KLH, keyhole limpet hemocyanin; SOCE, store-operated calcium entry; STIM1/STIM2, stromal interaction molecule 1/2; TCD, T cell-depleted bone marrow; Treg, regulatory T; wt, wild type; CBA, cytometric bead array; BM, bone marrow; TG, thapsigargin; Tind, indicator T cell; FSC, forward scatter; SSC, side scatter.
The online version of this article contains supplemental material.