Tumor suppressor TSC1 is critical for T-cell anergy

Abstract

T-cell anergy is a state of T cells that is hyporesponsive to stimulation via the T-cell receptor and costimulatory molecules and is thought to be important for self-tolerance. How T-cell anergy is regulated is still poorly understood. We report here that tuberous sclerosis (TSC)1 is critical for T-cell anergy. Deficiency of TSC1 resulted in enhanced T-cell proliferation and cytokine production in the absence of cluster of differentiation (CD)28-mediated costimulation, accompanied by enhanced T-cell metabolism. Resistance of TSC1-deficient T cells to anergy is correlated with increased signaling through the mammalian target of rapamycin complex (mTORC)1 and can be reverted by treatment of these cells with mTORC1 inhibitor rapamycin. Expression of the inducible costimulator (ICOS) is increased in TSC1-deficient T cells, which can be inhibited by rapamycin. Simultaneous blockade of both CD28 and ICOS costimulation partially restored sensitivity of TSC1-deficient T cells to anergy induction. Together, our data indicate that TSC1 is crucial for T-cell anergy by inhibiting mTORC1 signaling through both ICOS-dependent and -independent mechanisms.

Coordinated signals from the T-cell receptor (TCR), costimulatory receptors, and cytokine receptors ensure naïve T-cell activation and differentiation to proper effector T cells. T cells can be made anergic by stimulation through their TCR without a costimulatory signal, by stimulation with partial agonist peptides in the presence of costimulation, or by treatment with the Ca²⁺ ionophore ionomycin (1–3). Anergic T cells do not respond to antigen restimulation even in the presence of appropriate costimulation. They are impaired in producing cytokines such as IFN-γ or IL-2, defective in proliferative response, and metabolically inert (4–7). T-cell anergy is thought to be important for preventing self-reactive T cells from full activation to trigger autoimmune diseases.

The mammalian target of rapamycin (mTOR) integrates numerous environmental stimuli including growth factors, nutrients, and stress-activated signals to regulate cell metabolism, survival, growth, and proliferation (8). mTOR associates with multiple proteins and forms two distinct signaling complexes. mTOR complex (mTORC)1 contains raptor and phosphorylates ribosomal protein s6 kinase, 70-kd, 1 (S6K1) and eukaryotic translation initiation factor 4e-binding protein 1 (4E-BP1) to promote ribosomogenesis and protein translation. The rictor-containing mTORC2 phosphorylates both Akt at serine 473 residue to promote Akt activation and cell survival and PKCθ to promote T-helper (Th)2 differentiation (9, 10). Engagement of the TCR activates both mTORC1 and mTORC2, which is dependent on the RasGRP1-Ras-Erk1/2 pathway and is inhibited by diacylglycerol kinases (11–13). Rapamycin treatment or genetic ablation of mTOR induces T-cell anergy and inhibits helper T-cell differentiation but promotes expansion of natural regulatory T cells (nTregs) and the generation of inducible Tregs (9, 14–16). Contrary to its positive role in promoting T-cell activation and effector T-cell function during primary immune responses, mTOR inhibits memory CD8 T-cell formation and function during viral infection (17, 18). Moreover, mTOR regulates T-cell trafficking in vivo by modulating the expression of the chemokine receptor CCR7 (C-C chemokine receptor 7) (19). The importance of mTOR in T cells suggests that its activity must be tightly controlled.

The tuberous sclerosis (TSC)1/2, a heterodimer of TSC1 and TSC2, is a tumor suppressor that inhibits mTORC1. Activation of mTORC1 requires the association of GTP-bound active form of RheB (Ras homolog enriched in brain). TSC2 contains GAP activity for RheB and promotes the formation of GDP-bound inactive form of RheB. TSC1 does not have GAP activity, but its association with TSC2 prevents TSC2 from degradation (20). Several recent reports have demonstrated that TSC1 plays important roles for normal homeostasis of mitochondria in T cells, peripheral T-cell survival, and maintenance of T-cell quiescence (21–23). In this report, we demonstrate that TSC1 is critical for T-cell anergy by negative control of mTORC1. Its deficiency leads to cluster of differentiation (CD)28-independent activation of T cells that is manifested by increased cytokine production, proliferation, and metabolism. Resistance of TSC1-deficient T cells to anergy is dependent on mTORC1 activity and is correlated with enhanced up-regulation of inducible costimulator (ICOS). Blockade of ICOS signal partially restored anergy in TSC1-deficient T cells, suggesting that both ICOS-dependent and -independent mechanisms are involved in resistance of TSC1-deficient T cells to anergy.

Results

Deficiency of TSC1 Enhances T-Cell Proliferation in the Absence of CD28 Costimulation.

We first examined whether the expression of TSC1 is regulated during T-cell activation. We stimulated T cells with an anti-CD3 antibody in the presence of either anti-CD28 to induce full activation or cytotoxic T-lymphocyte antigen (CTLA)4-Ig to block CD28-mediated costimulation to induce anergy. TSC1 mRNA levels in sorted CD4⁺ and CD8⁺ T cells after 48-h treatment was compared with those in naïve CD44^lowCD62L⁺ CD4⁺ and CD8⁺ T cells. TSC1 mRNA was expressed at the highest level in naïve CD4⁺ T cells, down-regulated in activated CD4⁺ T cells, and remained at a high level in anergic CD4⁺ T cells (Fig. 1A). The expression of TSC1 in naïve CD8⁺ T cells was about 50% lower than CD4⁺ T cells. Although TSC1 expression was also down-regulated in activated and remained high in anergic CD8⁺ T cells, the differences were not as obvious as in CD4⁺ T cells.

Fig. 1.

TSC1 deficiency enhances T-cell proliferation in the absence of CD28-costimulation. (A) Expression of TSC1 in naïve, activated, and anergic T cells. T cells were activated with anti-CD3 and anti-CD28 or anergized with anti-CD3 plus CTLA4-Ig for 48 h. TSC1 mRNA in sorted naïve (CD44^lowCD62L⁺), activated, and anergized CD4 and CD8 T cells were measured by real-time qPCR and presented as mean ± SEM of triplicates in one experiment. Data shown are representative of two experiments. *P < 0.05, as determined by Student t test. (B) Enhanced proliferation of TSC1-deficient T cells under the anergy-inducing condition. Sorted WT and TSC1KO CD44⁻CD62L⁺ naïve CD4⁺ and CD8⁺ T cells were labeled with CFSE, mixed with mitomycin C–treated TCRβ^−/−δ^−/− splenocytes, and then either left unstimulated or stimulated with anti-CD3 plus anti-CD28 or CTLA4-Ig at 37 °C for 48 h. After staining for CD4 and CD8, cells were analyzed by flow cytometry. Overlaid histograms show CFSE intensities in gated WT and TSC1KO CD4 and CD8 T cells. Data shown are representative of three experiments.

To investigate whether TSC1 plays a role in T-cell activation, we bred TSC1^f/f mice (24) to CD4-Cre transgenic mice and generated TSC1^f/f-CD4-Cre mice (referred to as TSC1KO) to delete the TSC1 gene in T cells. In TSC1KO mice, T-cell development in the thymus is not grossly altered, but peripheral T-cell numbers are decreased because of increased T-cell death, correlating with an increase of the relative ratio of CD44⁺CD62L⁻ effector/memory to naïve T cells (21–23). To examine whether TSC1 regulates naïve T-cell proliferation, sorted WT and TSC1KO CD44⁻CD62L⁺ naïve CD4⁺ and CD8⁺ T cells were stimulated with anti-CD3 plus either anti-CD28 or CTLA4-Ig in the presence of antigen-presenting cells. When stimulated with anti-CD3 in the presence of anti-CD28, TSC1KO T cells proliferated similarly to WT T cells. When stimulated with anti-CD3 in the presence of CTLA4-Ig, proliferation of WT T cells was inhibited. However, TSC1KO T cells, particularly CD4⁺ T cells, proliferated vigorously (Fig. 1B). Thus, TSC1-deficient naïve T cells could be activated in the absence of CD28 costimulation.

TSC1-Deficient T Cells Are Resistant to Anergy.

The data described above indicated that TSC1 deficiency alleviated T cells from the requirement of CD28-mediated costimulation for activation. To further analyze whether TSC1 deficiency breaks T-cell anergy, we stimulated sorted naïve WT and TSC1KO CD4⁺ T cells in the presence or absence of anti-CD28 or CTLA4-Ig for 48 h. Following 24 h of resting, the cells were restimulated with anti-CD3 plus anti-CD28 to assess IL-2 and IFN-γ production. WT CD4⁺ T cells pretreated with anti-CD3 plus CTLA4-Ig contained much lower percentages of IL-2– and IFN-γ–producing cells than when they were pretreated with anti-CD3 plus anti-CD28, suggesting that they became anergic. However, TSC1KO CD4⁺ T cells pretreated with anti-CD3 plus CTLA4-Ig contained more IL-2– and IFN-γ–producing cells than similarly treated WT controls (Fig. 2 A and B). Moreover, TSC1KO CD4⁺ T cells appeared to produce higher levels of IL-2 and IFN-γ than WT controls under anergic condition, reflected by the increase in mean fluorescence intensity (MFI) of IL-2 or IFN-γ in producing TSC1KO CD4⁺ T cells that were IL-2⁺ or IFN-γ⁺. Thus, TSC1 deficiency conferred T-cell resistance to anergy induction in vitro.

Fig. 2.

TSC1 deficiency confers T-cell resistance to anergy induction. (A and B) TSC1KO T cells are resistant to anergy induction in vitro. Sorted WT and TSC1KO CD44⁻CD62L⁺ naïve CD4⁺ T cells were mixed with mitomycin C–treated TCRβ^−/−δ^−/− splenocytes, left unstimulated, or stimulated with anti-CD3 in the presence of anti-CD28 (Act) or CTLA4-Ig (Ane) at 37 °C for 48 h. After resting for 24 h, cells were restimulated with plate-bound anti-CD3 plus anti-CD28 for 12 or 24 h for intracellular staining of IL-2 and IFN-γ, respectively, and analyzed by flow cytometry. (A) Increased IL-2 production by TSC1KO CD4⁺ T cells pretreated with anergy-inducing condition. Dot plots show intracellular IL-2 staining in gated CD4⁺ T cells. The bar figure shows means ± SEM of relative percentages of IL-2–producing cells and relative MFI of IL-2 staining of IL-2⁺ cells from three experiments. (B) Increased IFN-γ production by TSC1KO CD4⁺ T cells pretreated with anergy-inducing condition. (C–E) TSC1 deficiency confers T-cell resistance to anergy induction in vivo. WT and TSC1KO mice received i.p. injections of 100 μg of SEB in 200 μL of PBS. Seven days after injection, splenocytes were stained for CD4, TCRVβ6, and TCRVβ8. (C) Bar graphs represent mean ± SEM of percentages of Vβ8⁺ or Vβ6⁺ cells in gated CD4⁺ T cells from one experiment (n = 5) that is representative of two experiments. (D) Splenocytes were stimulated with SEB for 12 or 24 h for intracellular staining of IL-2 and IFN-γ, respectively. (E) Splenocytes were also labeled with fluorescent dye eFluro 670 and stimulated with SEB for 72 h to assess proliferation. Histograms show eFluro 670 intensity in gated CD4⁺Vβ8⁺ and CD4⁺Vβ6⁺ cells. Data shown are representative of three (A and B) and two (C–E) experiments. *P < 0.05; **P < 0.01; ***P < 0.001, as determined by Student t test. (F) Mononuclear cell infiltration in the thyroid gland and liver of aged TSC1KO mice. Representative hematoxylin and eosin staining of thin sections are shown.

We further used the Staphylococcus enterotoxin (SE)B superantigen-induced T-cell anergy model to determine whether TSC1 deficiency affects T-cell anergy in vivo. SEB crosslinks Vβ8⁺ TCR and selectively induces anergy or deletion of Vβ8⁺ T cells (25, 26). WT and TSC1KO mice were administrated with SEB or PBS. One week later, splenocytes were collected and stimulated with SEB in vitro to examine IL-2 and IFN-γ production by intracellular staining and proliferation of Vβ8⁺ T cells using a fluorescent dye eFluro 670 dilution assay. As shown in Fig. 2C, the percentage of TCRVβ8⁺ T cells within the CD4⁺ T-cell population was similar between SEB-injected WT and TSC1KO mice, suggesting that SEB-induced deletion of the Vβ8⁺ T cells was not obviously affected by TSC1 deficiency. WT TCRVβ8⁺ CD4⁺ T cells from SEB-injected mice had fewer IL-2– and IFN-γ–producing cells than those WT cells from PBS-injected mice, suggesting effective induction of anergy by SEB injection (Fig. 2D). However, TSC1KO TCRVβ8⁺ CD4⁺ T cells contained similar percentages of IL-2– and IFN-γ–producing cells from mice injected with either SEB or PBS. Furthermore, TSC1KO TCRVβ8⁺ CD4⁺ T cells from SEB-injected mice proliferated much stronger than WT control following in vitro SEB stimulation (Fig. 2E). As expected, SEB did not induce proliferation of Vβ6⁺ CD4⁺ T cells from either WT or TSC1KO mice. Thus, TSC1 deficiency also confers T-cell resistance to anergy in vivo. Moreover, four out of five aged TSC1KO but not WT mice developed mild multiorgan mononuclear cell infiltration (Fig. 2F), suggesting that TSC1 is important for T-cell tolerance and for preventing autoimmunity.

Contribution of Enhanced mTORC1 Signaling to the Resistance of Anergy by TSC1-Deficient T Cells.

We further examined the effects of TSC1 deficiency on mTOR signaling in T cells under activation and anergy conditions. WT naïve CD4⁺ T cells under anergy-inducing conditions exhibit decreased 4E-BP1 phosphorylation compared with those WT T cells under activating conditions, indicating that mTORC1 signaling is defective in anergic T cells, which was consistent with previous observations (15). However, TSC1-deficient naïve CD4⁺ T cells displayed elevated 4E-BP1 phosphorylation in both activating and anergy-inducing conditions (Fig. 3A). Thus, TSC1 is critical for silencing mTORC1 activity in anergic T cells.

Fig. 3.

Contribution of enhanced mTORC1 signaling to the resistance to anergy of TSC1KO T cells. (A) Enhanced mTORC1 signaling in TSC1KO T cells. Sorted WT and TSC1KO CD44⁻CD62L⁺ naïve CD4⁺ T cells were similarly treated to induce activation and anergy as in Fig. 2A. T cells were purified after treatment, and cell lysates were subjected to immunoblotting analysis with the indicated antibodies. (B–E) Inhibition of IL-2 and IFN-γ production by WT and TSC1KO CD4⁺ T cells by rapamycin. WT and TSC1KO splenic T cells were stimulated with anti-CD3 in the presence of anti-CD28 (Act) or CTLA4-Ig (Ane) with or without 20 nM rapamycin at 37 °C for 48 h. After resting for 24 h, cells were restimulated with plate-bound anti-CD3 plus anti-CD28 for 12 or 24 h for intracellular staining of IL-2 and IFN-γ, respectively, and analyzed by flow cytometry. (B and D) Contour plot for IL-2 and IFN-γ expression in gated CD4⁺ T cells. (C and E) Mean ± SEM presentation of relative percentages of IL-2– and IFN-γ–producing WT and TSC1KO CD4⁺ T cells and relative IL-2 and IFN-γ MFI in IL-2⁺ or IFN-γ⁺ CD4⁺ T cells under an anergy-inducing condition from three experiments. *P < 0.05; **P < 0.01; ***P < 0.001, as determined by the Student t test.

To determine the role of enhanced mTORC1 activity in the resistance of TSC1KO T cells to anergy, we treated T cells with rapamycin during primary stimulation. Rapamycin treatment reduced not only the percentages but also cytokine levels of IL-2– and IFN-γ–producing cells in both WT and TSC1KO CD4⁺ T cells under activating condition. Under anergic conditions, rapamycin treatment was also able to significantly reduce IL-2 and IFN-γ expression in both WT and TSC1KO CD4⁺ T cells (Fig. 3 B–E). Thus, mTORC1 activity was important for T-cell activation, and increased mTORC1 activity contributed to resistance to anergy by TSC1KO T cells.

Enhanced T-Cell Metabolism in the Absence of TSC1.

During the transition from the naïve resting state to an activated state, T-cell metabolism is substantially increased, including increased uptake of nutrients, accelerated glycolysis, and massive protein and DNA synthesis. In contrast, anergic T cells are thought to be metabolically anergic (6). Because mTOR is a central regulator of metabolism and is regulated by TSC1, we investigated whether TSC1 controls T-cell metabolism. Naïve TSC1KO CD4⁺ T cells expressed similar levels of transferrin receptor protein 1 (CD71), which is required for iron delivery from transferrin to cell, to WT controls. Activated TSC1KO CD4⁺ T cells up-regulated slightly more CD71 than WT controls. CD71 up-regulation was stronger in TSC1KO CD4⁺ T cells than in WT control in anergizing condition (Fig. 4A). Expression of CD98, a heterodimeric membrane transport protein that preferentially transports branched-chain and aromatic amino acids, was also increased in TSC1KO naïve, activated, and anergized CD4⁺ T cells compared with WT controls (Fig. 4B). Glucose transporter 1 (Glut1) was expressed at higher levels in TSC1KO CD4⁺ T cells under either activating and anergizing conditions compared with WT controls (Fig. 4C). Consistent with increased Glut1 expression, glucose uptake was increased in TSC1KO T cells that had been activated or anergized compared with WT T cells treated in the same manner (Fig. 4D). Collectively, these data indicate that TSC1 negatively controls T-cell metabolism. The enhanced nutrient uptake and metabolism may be involved in the resistance of TSC1KO T cells to anergy.

Fig. 4.

TSC1 inhibits T-cell metabolism. (A and B) Increased CD71 and CD98 expression in TSC1KO T cells. Sorted WT and TSC1KO CD44⁻CD62L⁺ naïve CD4⁺ T cells were left unstimulated or stimulated with anti-CD3 plus anti-CD28 or plus CTLA4-Ig for 48 h as in Fig. 2A. Overlaid histograms show CD71 (A) and CD98 (B) expression in gated CD4⁺ T cells. (C) Increased Glut1 mRNA levels in TSC1KO CD4⁺ T cells following activating or anergizing treatment. (D) Enhanced glucose uptake by TSC1KO T cells. WT and TSC1KO T cells following primary activating or anergizing stimulation, resting, and anti-CD3 plus anti-CD28 restimulation were subjected to glucose uptake assay. Data shown are representative of three experiments. *P < 0.05; **P < 0.01; ***P < 0.001, as determined by the Student t test.

Elevated ICOS Expression Partially Contributes to Resistance of Anergy by TSC1KO T Cells.

TCR engagement without costimulation promotes T-cell anergy, in part, by inducing early growth response (Egr)2/3 and E3 ubiquitin ligases itchy E3 ubiquitin protein ligase (Itch), gene related to anergy in lymphocytes (Grail), and Casitas B-lineage lymphoma b (Cbl-b) through the calcium–nuclear factor of activated T cells (NFAT) pathway (27–29). Expression of Egr2/3 and Itch, Grail, and Cbl-b was decreased in TSC1KO CD4⁺ T cells treated with anti-CD3 plus CTLA-4-Ig compared with WT control (Fig. 5 A and B). Diacylglycerol kinases (DGK) α and ζ terminate diacylglycerol (DAG)-mediated signaling to ensure the dominance of calcium signaling in anergic T cells (25, 27, 30). In TSC1KO CD4⁺ T cells under anergizing treatment, DGKζ but not DGKα expression was decreased (Fig. 5A). Cbl-b targets phospholipase (PL)Cγ1 and PKCθ, whereas Grail may target CD3ζ for degradation (31). CD3ζ protein was increased in both activated and anergized TSC1KO CD4⁺ T cells compared with WT control. Although PLCγ1 protein was not increased and PKCθ protein was actually decreased in anergized TSC1KO CD4⁺ T cells, their phosphorylation was increased, suggesting enhanced signaling via these molecules.

Fig. 5.

Elevated ICOS expression contributes to CD28-independent activation of TSC1KO T cells. Sorted naïve WT and TSC1KO CD4⁺ T cells were induced for activation and anergy similarly to Fig. 2A or directly used to make RNA. Forty-eight hours later, live CD4⁺ T cells were sorted for making RNA and total cell lysates. mRNA levels of indicated molecules were quantified by real-time qPCR. (A) Bar graphs are means ± SEM of triplicates from one experiment and represent two experiments. (B) Protein phosphorylation and total protein levels were determined by immunoblotting analysis with the indicated antibodies. Data shown represent two experiments. (C) Cell surface ICOS protein expression on gated CD4⁺ T cells at the indicated times were assessed by FACS analysis. Data shown represent three experiments. (D) Blockage of ICOS signal partially restored TSC1KO T-cell anergy. Sorted naïve WT and TSC1KO CD4⁺ T cells were induced for activation or anergy similar to Fig. 2 except that anti-ICOSL or control Ig was added during the 48 h treatment. Bar figure shows means ± SEM of relative percentages of IFN-γ–producing cells calculated from three experiments.

ICOS provides an inducible costimulatory signal for T cells. Although ICOS expression was similar between naive WT and TSC1KO CD4⁺ T cells (Fig. 5 A and C), ICOS mRNA was increased in TSC1KO CD4⁺ T cells under both activating and anergic condition. Under activating condition, WT CD4⁺ T cells gradually up-regulated surface ICOS, which reached the highest level at 48 h and decreased almost to the basal level following an additional 24-h resting. TSC1 deficiency appeared to accelerate cell surface ICOS up-regulation and maintained ICOS levels following 24-h resting. Under anergizing condition, cell surface ICOS expression in TSC1KO CD4⁺ T cells was similar to or slightly higher than WT CD4⁺ T cells at 3, 12, and 48 h during anergy induction. However, TSC1KO CD4⁺ T cells maintained ICOS expression at a higher level than WT CD4⁺ T cells upon resting (Fig. 5C). Correlated with increased ICOS expression, Roquin, a protein posttranscriptionally repressing ICOS expression (32, 33), was decreased in TSC1KO CD4⁺ T cells (Fig. 5A).

To test whether increased ICOS expression may confer TSC1KO T cells CD28-independent activation, we treated naïve WT and TSC1KO CD4⁺ T cells with anti-CD3 plus CTLA4-Ig in the presence or absence of an anti-ICOSL antibody. Blockade of ICOS signal by anti-ICOSL decreased IFN-γ production in both WT and TSC1KO CD4⁺ T cells under either activating or anergizing conditions (Fig. 5D), suggesting that elevated ICOS signaling may contribute to TSC1KO CD4⁺ T-cell activation in the absence of CD28-mediated costimulatory signal. Because TSC1KO CD4⁺ T cells still produced slightly higher levels of IFN-γ than WT control under anergic condition when ICOS signal was blocked, it suggests that ICOS-independent mechanisms may also contribute to resistance to anergy by TSC1KO T cells.

Discussion

The mechanisms that control T-cell anergy are not fully understood. Using both in vitro and in vivo T-cell anergy model systems, we have demonstrated that TSC1 deficiency results in CD28-independent T-cell activation that is manifested by increased nutrient uptake, enhanced proliferation, and elevated cytokine production. TSC1 promotes T-cell anergy by inhibiting mTORC1 signaling.

In the absence of TSC1, ICOS is up-regulated in T cells, which is partially responsible for the resistance of TSC1-deficient T cells to anergy. It is interesting to note that both B7.1/7.2-CD28 and B7h-ICOS costimulatory signals function in an inducible manner. The B7.1/7.2-CD28 signal is induced through the induction of ligands, whereas CD28 is constitutively expressed on naïve T cells. Immature dendritic cells express very low levels of B7.1/7.2. B7.1/7.2 are up-regulated during dendritic maturation usually following microbial infection (34–36). In contrast, the B7h-ICOS signal is induced through ICOS up-regulation in T cells, whereas B7h is constitutively expressed on antigen-presenting cells. ICOS is expressed at a very low level in naïve T cells but is up-regulated during T-cell activation (37). Up-regulation of ICOS is promoted by CD28 and is dependent on mTOR signaling. Engagement of ICOS with its ligand B7h provides additional costimulation to promote full T-cell activation (38). Absence of TSC1 may partially relieve the requirement of CD28 for ICOS up-regulation via accelerating ICOS up-regulation. ICOS, when expressed at high levels, may be activated by its ligand B7h and substitute for CD28 to prevent T-cell anergy. Elevated ICOS expression, thus, may provide costimulation of T cells in the absence of microbial infection to prevent anergy. Both transcriptional and posttranscriptional mechanisms regulate ICOS expression (32). Because ICOS mRNA is substantially increased in TSC1KO T cells, TSC1 may control ICOS expression at least via transcriptional regulation.

Multiple mechanisms have been identified as being important for T-cell anergy (4). T-cell anergy can be caused by defects in proximal TCR signaling events (39), impaired Ras-Erk1/2 signaling (40, 41), selective activation of the Ca²⁺-NFAT pathway, and expression of anergy-promoting molecules such as Fas ligand (FasL) (42), transcription factors such as Erg2/3 (43), cell cycle inhibitors such as p27^kip1 (44), DGKα and -ζ (25, 27, 30), and E3 ubiquitin ligases (Cbl-b, Itch, and Grail) (28). In addition, epigenetic silencing of the IL-2 locus may contribute to the impaired IL-2 production in anergic T cells (45). In TSC1KO T cells, Egr2/3, Itch, Grail, and Cbl-b expression is down-regulated compared with WT T cells following induction of anergy, suggesting that TSC1 may promote T-cell anergy through up-regulation of these anergy-inducing molecules. Future studies are needed to determine how TSC1 controls the expression of these molecules. Furthermore, a recent report has demonstrated that Cbl-b prevents T-cell anergy, in part, by suppressing TCR/CD28-induced inactivation of phosphatase and tensin homolog (Pten) to prevent Akt activation (46). Because Akt promotes mTORC1 signaling by inactivating TSC2, it would be interesting to determine whether mTOR signaling is enhanced in Cbl-b–deficient T cells and whether enhanced mTOR signaling may contribute the loss of T-cell tolerance caused by Cbl-b deficiency.

In TSC1-deficient mice, particularly in aged mice, CD44⁺ effector T cells are increased compared with WT mice. Several mechanisms may contribute to the increase of CD44⁺CD62L⁻ effector T cells in these mice. The decreased peripheral T-cell number in TSC1-deficient mice may result in increased homeostatic proliferation of T cells, which is accompanied by CD44 up-regulation. TSC1 has also been found to be important for T-cell quiescence, and loss of quiescence could lead to an activated phenotype of TSC1-deficient T cells (22, 23). The data presented in this report reveal that TSC1 is critical for T-cell anergy in vitro and in vivo. It is well known that negative selection during intrathymic T-cell development is not perfect; some self-reactive T cells escape negative selection and are part of the normal T-cell repertoire. These self-reactive T cells are prevented from full activation by self-antigens via Treg-mediated suppression and by induction of anergy. The resistance of TSC1-deficient T cells to anergy may also lead to activation and accumulation of self-reactive T cells and increase of CD44⁺CD62L⁻ T cells and the development of autoimmune diseases in TSC1-deficient mice.

Materials and Methods

Mice and Cells.

The TSC1^flox/flox, TCRβ^−/−δ^−/−, and CD4-Cre transgenic mice were purchased from The Jackson Laboratory and Taconic Farm, respectively (24). All experiments were performed in accordance with protocols approved by the Institutional Animal Care and Use Committee at Duke University. Single-cell suspension of splenocytes and lymph node (LN) cells were made in Iscove's Modified Dulbecco's Medium (IMDM) with 10% (vol/vol) FBS, penicillin/streptomycin, and 2-mercaptoethanol (IMDM-10). Purification of T cells was achieved using either the Mouse T Cell Enrichment Kit (STEMCELL Technologies) or the LS depletion columns (Miltenyi Biotech), and purities of ≥90% were achieved. When indicated, purified T cells were further stained for CD44 and CD62L and sorted for naïve T cells using a MoFlo sorter with purities >98%.

Flow Cytometry.

Cells were stained with fluorescence conjugated antibodies in 2% (vol/vol) FBS in PBS at 4 °C for 30 min. Cell death was identified by using the Violet Live/Dead cell kit (Invitrogen). The Cytofix/CytopermFixation/Permeabilization kit (BD Bioscience) was used for intracellular staining. Data were collected using a BD Canto II and analyzed with FlowJo (Treestar). For comparison purposes, relative percentages of IFN-γ– and IL-2–producing cells within the CD4⁺ T cells among different experiments was normalized to percentages of cytokine-producing WT T cells under the activating condition (set as 100) in each experiment. Relative MFI of IFN-γ⁺ or IL2⁺ cells were similarly normalized to WT T cells producing these cytokines.

T-Cell Proliferation and in Vitro Anergy Assay.

For the primary stimulation, sorted naïve CD44⁻CD62L⁺ WT and TSC1KO-T splenocytes and LN cells were either unlabeled or labeled with carboxyfluorescein diacetate succinimidyl ester (CFSE) as described previously (47). Cells were seeded in a 24-well plate (3-5 × 10⁶) or 48-well plate (1.0–2.5 × 10⁶) with mitomycin C–treated TCRβ^−/−δ^−/− splenocytes in 1 mL of IMDM-10, left unstimulated, or stimulated with anti-CD3 (1.0 μg/mL; 2C-11) plus anti-CD28 (0.5 μg/mL; 37.51) or CTLA4-Ig (10 μg/mL; BioXcell) at 37 °C. As indicated, rapamycin (20 nM) or anti-ICOSL (60 μg/mL; BioXcell) was added during primary stimulation. Forty-eight hours later, CFSE-labeled cells were stained for CD4 and CD8 to assess proliferation. Cells not labeled with CFSE were washed for three times and rested in IMDM-10 at 37 °C for 24 h. Live cells enriched after Lympholyte (CEDARLANE) gradient separation were restimulated with plate-bound anti-CD3 (1 μg/mL) and soluble anti-CD28 (0.5 μg/mL) in the presence of 5 μM monensin at 37 °C for 12 or 24 h for intracellular staining of IL-2 and IFN-γ, respectively.

SEB in Vivo Anergy.

SEB-induced anergy was examined as reported previously with modifications (25, 26). Briefly, WT and TSC1KO mice received i.p. injections of 100 μg of SEB (Sigma) in 200 μL of PBS. Seven days later, splenocytes were used for staining or were seeded at 5 × 10⁵ per well in a U-bottomed 96-well plate and left unstimulated or stimulated with 10 μg/mL SEB in the presence of Golgi Plug at 37 °C for 24 h, followed by cell surface staining for CD4 and TCRVβ8 and intracellular staining for IFN-γ and IL-2. Splenocytes were also labeled with eFluro 670 (2.5 μM at room temperature for 9 min according to the protocol of the manufacturer; eBioscience) and stimulated with SEB at 37 °C for 72 h, followed by staining for CD4, TCRVβ6, and -8 and FACS analysis.

Histology.

Liver and thyroid gland from 1-y-old WT and TSC1KO mice were frozen in OCT medium, and thin sections were made and stained with hematoxylin and eosin following standard procedures.

Immunoblot.

T cells treated with activating and anergizing conditions were enriched with anti-CD90.2 magnetic beads and the LS depletion columns with about 90% purities. Purified T-cell lysates were subjected to immunoblotting analysis as described previously (11). Anti–Cbl-b, anti-PKCθ, and anti-CD3ζ were from Santa Cruz Biotechnology. Anti–phospho-PKCθ/δ, anti–phospho-PLCγ1, anti–phosphor-S6K, and anti-S6K were from Cell Signaling Technology.

Real-Time PCR.

RNA was isolated from sorted T cells using TRIzol reagent (Sigma). cDNA was synthesized using SuperScript III and random primers, and quantitative (q)RT-PCR was performed as described previously (48). Expressed levels of target mRNAs were normalized with β-actin and calculated using the 2^–ΔΔCT method. Primers include TSC1(forward, 5′-ACATCTTTGGCCGTCTCTCGTCAT-3′; reverse, 5′-ATGGGTACATCCCATAAAGGCGGT-3′), ICOS (forward, 5′-GGCAGACATGAAGCCGTACT-3′; reverse, 5′-CCTCCATTGTGAAATGAAAACA-3′), Glut1 (forward, 5′-GTATCCTGTTGCCCTTCTGC-3′; reverse, 5′-TGTCCCTCGAAGCTTCTTCA-3′), Egr2 (forward, 5′-TTGACCAGATGAACGGAGTG-3′; reverse, 5′-CAGAGATGGGAGCGAAGCTA-3′), Egr3 (forward, 5′-CGCGCTCAACCTCTTCTC-3′; reverse, 5′-GATTGGGCTTCTCGTTGGT-3′); Itch (forward, 5′-GGACCACAGCTTGATTCCAT-3′; reverse, 5′-GACTTGGTCCAAACCAATTCTT-3′), Grail (forward, 5′-CAACCGTGGGCTATTTCATC-3′; reverse, 5′-GCAGCTGAAGCTTTCCAATA-3′), DGKα (forward, 5′-CCAGAGAAGGAAGCGTTGAC-3′; reverse, 5′-GCTTCTTGTTTTTCAGAGGGTTA-3′), DGKζ (forward, 5′-CTGAGGAGCAGATCCAGAGC-3′; reverse, 5′-TCCCCGACATAGCAGAAGTC-3′), and Roquin (forward, 5′-TGAGGAAGCCAAGAAATGTG-3′; reverse, 5′-ATTGGGCGACTCAGAACACT-3′).

Glucose Uptake.

Following primary stimulation, resting, and restimulation for 24 h as described, live T cells were enriched by Lympholyte, followed by anti-CD90.2 magnetic beads and MACS purification. Glucose uptake by purified T cells (0.5 × 10⁶/mL cells for each reaction) was performed according to a published protocol (49).

Statistical Analysis.

Statistical significance was determined using the Student t test. P values are defined as follows: P < 0.05, P < 0.01, and P < 0.001.