Interleukin 2 gene transcription is regulated by Ikaros-induced changes in histone acetylation in anergic T cells

Random generation of antigen receptors allows the adaptive immune system to initiate specific and effective responses against a vast array of invading pathogens. Randomly rearranged receptors, however, also have the potential to recognize self-antigens. To prevent the catastrophic consequences of an immune attack against the body's own constituents, several mechanisms are in place to control the activity of self-reactive lymphocytes. Most T cells that carry self-reactive T-cell receptors (TCRs) are clonally eliminated in the thymus; however, many of these cells survive this process and must be inactivated in the periphery. One of the mechanisms that contributes to the establishment of peripheral T-cell tolerance is anergy: an intrinsic process that leads to the inactivation of self-reactive T cells. In anergic T cells, the ability to proliferate and synthesize interleukin 2 (IL-2) upon antigen encounter is profoundly reduced, whereas the expression of other cytokines is affected to different degrees.

Signals transmitted through the TCR are able to implement 2 opposite programs: activation and anergy. The presence or absence of costimulatory and/or negative signals will determine the outcome. Thus, TCR engagement together with signals transmitted by CD28 induces T-cell activation, whereas in the absence of such costimulation, TCR engagement will lead to anergy.^1-4  Negative signals derived from receptors such as CTLA-4 also contribute to the establishment of T-cell anergy in vivo.^5,6 

One of the consequences of TCR engagement in the absence of costimulation is a sustained increase in intracellular calcium, which leads to the activation of the calmodulin-dependent phosphatase calcineurin (Cn).⁷  Activated Cn dephosphorylates and induces the nuclear translocation of nuclear factor of activated T cells (NFAT) proteins, a family of transcription factors that play crucial roles in T-cell development and function.^8,9  Members of the AP-1 family of transcription factors are the main transcriptional copartners of NFAT in T cells and require costimulation to be fully activated.¹⁰  In the absence of AP-1 cooperation, NFAT proteins induce the expression of a set of genes that are specifically up-regulated in T cells in response to anergizing stimuli.⁷  The mechanisms that induce unresponsiveness in anergic T cells are currently being elucidated. Recently, several ubiquitin ligases, whose expression is induced by calcium/Cn, have been shown to participate in the blocking of TCR signaling in anergic T cells by modifying or targeting components of the TCR signalosome for degradation.^11-14  The analysis of specific genes activated in anergic T cells supports the existence of different mechanisms of tolerance induction in T lymphocytes, including not only protein degradation but also interference with signaling pathways coupled to antigen receptors, direct control of cell-cycle progression, and transcriptional modulation.⁷ 

Covalent modifications of the chromatin in specific regions of the genome can regulate the interactions of different proteins with the DNA, controlling processes such as recombination, replication, or gene transcription. Active chromatin generally associates with hyperacetylated histones, whereas silenced or inactive chromatin is usually associated with deacetylated histones. These modifications are the result of the activity of histone acetyl-transferases and histone deacetylases (HDACs). Targeting of these enzymes to a specific locus is likely mediated by transcriptional activators or repressors that recognize specific DNA motifs and are able to recruit them.^15,16  The expression of Ikaros, the founding member of a family of transcription factors with essential roles in lymphoid lineage development,^17-19  is up-regulated during the induction of T-cell anergy.⁷  Ikaros has a well-characterized transcriptional repressor activity mediated by its ability to recruit chromatin remodeling complexes containing HDACs.²⁰ 

In this study we intended to determine whether epigenetic changes might regulate the expression of IL-2 in anergic cells. Our results indicate that during anergy induction, calcium signaling induces histone deacetylation of the il2 promoter in T-helper 1 (Th1) cells. We show that anergizing stimuli induce up-regulation of the expression of Ikaros, which binds to the il2 promoter and induces deacetylation of histone 4 (H4) and silencing of il2 expression. Furthermore, we present evidence that this direct mechanism of transcriptional repression may be crucial to achieve a complete inhibition of il2 expression in anergic T cells.

Therefore, we propose a model in which signals transmitted by calcium in response to anergic stimuli induce the expression of transcriptional repressors that cause stable epigenetic changes that lead to silencing of the il2 gene. These mechanisms play a critical role in the maintenance of anergic T-cell unresponsiveness.

Wild-type C57BL6/J or DO11.10 mice were purchased from Jackson Laboratories (Bar Harbor, ME) and were maintained in pathogen-free conditions. All animal work was performed according to the guidelines set by the Albert Einstein College of Medicine Institutional Animal Care and Use Committee.

CD4⁺ T cells were isolated from lymph nodes and spleens from 4- to 6-week-old mice using Dynabeads-CD4 magnetic beads (Dynal, Lake Success, NY). CD4⁺ T cells were stimulated with plate-bound anti–mouse CD3ϵ and CD28 (BD, San Jose, CA) at 0.5 μg/mL. Th1 cells were differentiated for 7 days in the presence of mouse IL-12 (10 ng/mL), anti–mouse IL-4 (10 μg/mL), and mouse IL-2 (10 units/mL; BD). T cells were cultured in RPMI. Jurkat and Phoenix Ecotropic (a gift from G. P. Nolan) cells were cultured in DMEM media supplemented with 10% FBS, 2 mM L-glutamine, and 50 μM β-mercaptoethanol.

A single dose of 25 mg of ovalbumin (Sigma, St Louis, MO; 100 mg/mL in PBS) was orally administered to DO11.10 mice. Seven days later, CD4⁺ T cells were isolated from these mice and from age- and sex-matched PBS-placebo–fed controls.

Th1 cells were treated with 1 μM ionomycin (Calbiochem, San Diego, CA) for 16 hours. In some experiments, trichostatin A (TSA; Upstate Biotechnology, Lake Placid, NY) at 10 nM was also added during the ionomycin treatment. Cells were then washed and rested for 2 to 4 hours in fresh medium before stimulation. Alternatively, cells were stimulated for 16 to 20 hours with 1 to 2 μg/mL plate-bound anti-CD3. Cells were then detached from the wells, washed, and allowed to rest for 72 hours before analysis.

Cells were stimulated for 4 hours with anti-CD3 and anti-CD28. For the last 2 hours, Brefeldin A (Sigma-Aldrich, St Louis, MO) was added to the culture at 10 μg/mL. Cells were then fixed in 4% paraformaldehyde, permeabilized in 0.5% saponin, and incubated in with 10 μg/mL phycoerythrin (PE)–conjugated anti–mouse IL-2 antibody (BD) for 30 minutes. Washed cells were analyzed on a FACSCAN analyzer (Becton Dickinson, San Jose, CA).

Culture supernatants were collected 24 hours following T-cell stimulation and IL-2 levels were measured in a sandwich enzyme-linked immunosorbent assay (ELISA; BD) following the manufacturer's recommendations.

Primary Th1 cells were stimulated with anti-CD3 and anti-CD28 for 48 hours and then labeled with 5-bromo-2′-deoxy-uridine for 12 hours. Incorporation of 5-bromo-2′-deoxy-uridine was determined using a BrdU labeling and detection kit (Roche, Indianapolis, IN) according to the manufacturer's instructions.

Jurkat cells were transfected by electroporation in serum-free medium with pulses of 250 V and 960 μF. Plasmids used include pIL-2 (−585)–Luc (5 μg per transfection), containing the murine il2 promoter (kindly provided by D. J. McKean), and the Ikaros1-expressing plasmid pCDM8-Ik-1 (a gift from K. Georgopoulos). Twenty-four hours after transfection, cells were stimulated with 500 nM ionomycin and 20 nM phorbol-12-myristate-13-acetate (PMA; Calbiochem) or plate-bound anti–human CD3 and CD28 (BD) at 1 μg/mL. Eight hours after stimulation, cells were lysed and assayed for luciferase activity using a Dual-Luciferase Reporter Assay System (Promega, Madison, WI). Cotransfection of a Renilla-luciferase plasmid was performed to normalize assays.

The retroviral vector RV-IRES-GFP has been previously described.⁷  The RV-HA-Ikaros1-IRES-GFP contains a cDNA encoding murine Ikaros1 HA-tagged at its 5′ end. Phoenix Ecotropic packaging cells were transfected using a calcium/phosphate method with retroviral vectors. Retroviral supernatants were collected 48 hours after transfection, supplemented with polybrene (8 μg/mL), and used to infect CD4⁺ T cells 24 and 48 hours after stimulation. Infected cells were sorted for GFP expression. Ikaros expression was assessed by Western blot using anti-Ikaros (Santa Cruz Biotechnology, Santa Cruz, CA; or a gift from S. T. Smale) or anti-HA (Zymed, South San Francisco, CA) antibodies.

Histone acetylation status and Ikaros binding were assayed using a ChIP assay kit (Upstate Biotechnology), following the manufacturer's protocol. Nuclear lysates from 2 × 10⁶ to 10 × 10⁶ cells were subjected to immunoprecipitation overnight at 4°C with antiacetylated H4, anti-H4 (Upstate Biotechnology), anti-Ikaros, or anti-HA antibodies. Specific primer pairs were designed to amplify the il2 promoter (5′-TTAAAACAACAACGACAA AATAG/5′-GTGGTGGACAAACGCAAGAG), the CD3ϵ promoter (5′-CATTTCCAAGTGACGTGG/5′-AACACACTGGCTGCATGC), and the murine β-actin gene (5′-AAGGACTCCTATGTGGGTGACGA/5′-ATCTTCTCCATGTCGTCCCAGTTC). Polymerase chain reaction (PCR) products were visualized on an agarose gel stained with ethidium bromide. Band intensities were quantified using Image J software (National Institutes of Health [NIH], Bethesda, MD). For quantification, the intensity of any given band was corrected by the amount of input, and background or isotype control intensities were subtracted.

Reactions were performed in a SmartCycler (Cepheid, Sunnyvale, CA) using SYBR Green PCR mix (Abgene, Rochester, NY) and primers for mouse Ikaros (5′-GCTGGCTCTCAAGGAGGAG, 5′-CGCACTTGTACACCTTCAGC). A threshold was set in the linear range of the amplification curve (fluorescence = f[cycle]), and the number of cycles needed to reach it (Ct) was calculated for every sample. Fold induction was calculated as 2^−ΔΔCt, using primers for actin as control. Melting curves and agarose gel electrophoresis established the specificity of the amplified band.

Ik-19i, an Ikaros-specific siRNA (5′-AAGACCTGTGCAAGATAGGAT-3′), was introduced into the pSUPER RNAi (Oligoengine, Seattle, WA). Control and pSUPER-IK-19i constructs were used together with a GFP plasmid to transfect Th1 cells by electroporation using the Mouse T-cell Nucleofector kit (AMAXA, Koeln, Germany) according to the manufacturer's protocol. Thirty-six hours after transfection, cells were sorted for GFP expression and then anergized with ionomycin. IL-2 production was assayed by ELISA following stimulation.

In order to determine how histone acetylation changes correlated with T-cell activation and IL-2 expression, we compared H4 acetylation status in naive and in resting or activated Th1 cells. In accordance with previous reports,²¹  naive CD4⁺ T cells showed a low level of H4 acetylation, which correlates with their limited ability to express IL-2. Upon differentiation, and having therefore gained increased capacity to transcribe the il2 promoter, Th1 cells showed enhanced H4 acetylation that was further increased after stimulation (Figure 1A). To determine whether this effect was specific for the IL-2 locus, we used primers to amplify the CD3ϵ promoter in samples obtained from naive and from resting and stimulated Th1 cells. A slight increase of H4 acetylation was observed in Th1 cells compared with naive cells. Nevertheless, stimulation did not induce further increase of H4 acetylation on the CD3ϵ promoter (Figure 1B).

Anergic T cells are characterized by their inability to proliferate and produce IL-2 in response to TCR stimulation. We had previously shown that calcium-mediated signals were necessary for the induction of clonal anergy in Th1 cells.⁷  To determine whether unopposed calcium signaling could not only result in defects in TCR signaling but also exert a direct effect on the regulation of IL-2 expression, we used a model of clonal anergy in mouse Th1 cells by inducing a sustained increase in intracellular calcium with ionomycin. Ionomycin-treated Th1 cells were unresponsive and unable to produce IL-2 when restimulated (Figure 2A). A different model of clonal anergy induction was also analyzed by stimulating Th1 cells with anti-CD3 (signal1) in the absence of costimulation (signal 2). Unresponsiveness to TCR engagement in these cells was also assessed by failure to proliferate or produce IL-2 in response to full stimulation (Figure 2B). In anergic T cells, these defects reflect an inability to activate il2 transcription.⁷  To determine whether direct chromatin modifications of the il2 promoter could underlie and contribute to this defect in IL-2 expression, we analyzed H4 acetylation by ChIP using antiacetylated H4 antibodies. As shown in Figure 1, in Th1 cells, which had been primed 1 week before analysis and had gained the capacity to express IL-2, H4 molecules on the il2 promoter were clearly acetylated (Figure 1A and Figure 2C, Ctrl lane). However, when those same cells were subjected to an anergizing stimulus, a drastic loss of H4 acetylation was clearly detected (Figure 2C). H4 acetylation was reduced to levels even below those detected in naive T cells. Histone deacetylation was prevented when TSA was added during anergy induction to inhibit HDAC activity (Figure 2C). To rule out the possibility that the absence of detectable amplified ChIP fragments in the anergic samples could be the result of nucleosome mobilization, we performed ChIPs using an antibody that recognized H4 independently of its acetylation status. Similar amounts of chromatin from the il2 promoter were precipitated in control and anergic cells with this antibody (Figure 2D). The specificity of this effect was determined using primers for the CD3ϵ promoter, whose expression is not affected in anergic T cells. No differences in H4 acetylation were detected between control and anergized cells on this locus (Figure 2E). Anergic T cells present a long-lasting status of unresponsiveness to antigen stimulation. Epigenetic changes in the il2 promoter could account for the imposition of a stable closed inactive status on this locus. To test this possibility, Th1 cells were anergized by stimulation with anti-CD3 and left resting for 3 days before performing ChIP assays to assess the level of H4 acetylation. These experiments showed that loss of H4 acetylation at the il2 promoter was stable in anergic Th1 cells (Figure 2F). As shown in the experiments performed with ionomycin, TSA prevented anti-CD3–induced H4 deacetylation of the il2 promoter but had no effect on control untreated Th1 cells (Figure 2F). We had previously seen that increased IL-2 expression upon stimulation was associated with increased H4 acetylation of the IL-2 promoter (Figures 1A, 2G). We wanted to determine not only whether the il2 promoter in anergic T cells underwent H4 active histone deacetylation but also whether the inability to synthesize IL-2 in response to TCR engagement also correlated with an inability to increase histone acetylation at this locus. ChIP assays using antiacetylated H4 were carried out in previously anergized Th1 cells restimulated with anti-CD3 and anti-CD28. As opposed to control cells, anergic T cells could not induce enhanced histone acetylation of the IL-2 promoter upon stimulation (Figure 2G).

To evaluate the contribution of anergy-induced histone deacetylation to the overall defect in IL-2 synthesis in anergic T cells, we inhibited anergy-induced HDAC activity using TSA. Th1 cells were treated with ionomycin in the presence or absence of TSA, and IL-2 production in response to TCR stimulation was analyzed. As expected, ionomycin induced an unresponsive status that was characterized by a profound defect in IL-2 production following stimulation (Figure 3A). TSA had no effect on the ability of control nonanergic T cells to make IL-2; however, TSA partially reversed the anergizing effect of ionomycin (Figure 3A left panel). Cells anergized in the presence of TSA were able to produce 3 times as much IL-2 as control anergic cells (P < .05 when comparing means with a t test for paired samples). These results suggested that TCR signaling blockade, although playing an important role in preventing IL-2 expression in anergic T cells, needed complementary mechanisms of active silencing to completely shut down IL-2 production. In order to bypass proximal blocks on TCR signaling in anergic T cells, we evaluated the response of these cells to stimulation with PMA and ionomycin. As previously shown,^22,23  cells pretreated with ionomycin showed only partial recovery when activated with PMA plus ionomycin, which suggested the presence of other blocks downstream of protein kinase C. Recovery was, however, almost complete when TSA was added during anergy induction (Figure 3A right panel). Inhibition of HDAC activity with TSA in anergic cells during a secondary stimulation was also able to partially restore IL-2 expression (Figure 3B), suggesting that mechanisms to prevent reacetylation—most likely by inducing further deacetylation of histones—were engaged in anergic T cells when restimulated. The presence of TSA during restimulation had, however, no effect on control cells or cells anergized in the presence of TSA (Figure 3B; data not shown).

We had previously shown that the expression of a specific set of genes was up-regulated during anergy induction in T cells. Supporting a possible role of Ikaros in T-cell anergy, we were able to detect a clear up-regulation of Ikaros mRNA expression in Th1 cells in response to calcium signals (3- to 4-fold) that translated into an increased accumulation of Ikaros protein in those cells (Figure 4A).

Sequence analysis had revealed that the il2 promoter might contain several binding sites for Ikaros.²⁴  To determine whether Ikaros could bind the il2 promoter in anergic T cells, we performed ChIP assays using anti-Ikaros antibodies. These experiments showed direct Ikaros binding to the il2 promoter in anergic Th1 cells (Figure 4B). We were also able to detect similar Ikaros binding to the IL-2 promoter in an in vivo model of T-cell tolerance. Oral administration of antigens has been shown to induce peripheral T-cell tolerance by 2 different mechanisms: deletion/anergy or induction of regulatory T cells. The specific outcome is determined by the administration regimen. High doses, like the one used in our experiments, preferentially induce anergy, whereas low doses induce a new population of T cells identified as regulatory T cells.²⁵  We analyzed CD4⁺ T cells from DO11.10 mice, which express a transgenic TCR that recognizes ovalbumin_323-339, fed with high doses of ovalbumin and confirmed they were anergic (Figure 4C). We had previously seen that these T cells up-regulated the expression of several potential anergy-inducing genes, including Ikaros.⁷  When ChIPs were performed on these cells using an antibody against Ikaros, a clear increase in Ikaros binding to the il2 promoter (2-fold) was detected in cells from orally tolerized mice compared with the control PBS-fed mice (Figure 4C).

Binding of Ikaros to the il2 promoter proved to be specific, as no amplified fragments could be detected when primers for mouse actin were used under the same conditions in which clear binding was detected using il2 promoter primers (compare Figure 4B and 4D). To determine whether HDAC activity was also being recruited together with Ikaros to the il2 promoter in response to unopposed calcium signaling, ChIP assays were performed using an anti-HDAC2 antibody. HDAC2 recruitment to the il2 promoter could only be detected in Th1 cells anergized with ionomycin but not in control untreated Th1 cells or cells anergized in the presence of the HDAC inhibitor TSA (Figure 4E).

In order to determine whether Ikaros could repress il2 expression, we transfected Jurkat cells with a luciferase reporter plasmid containing the murine il2 promoter and increasing amounts of an Ikaros-expressing plasmid. Under these conditions, transcription of the il2 promoter was inhibited by Ikaros in a dose-dependent manner (Figure 5A). Using retroviral vectors allowed us to carry out our studies in primary T cells and avoid problems associated with overexpression. Levels of Ikaros expression attained in retrovirally infected Th1 cells (RV-HA-Ikaros-IRES-GFP virus) were very similar to those detected in response to calcium stimulation (Figure 5B). In transduced cells, hemagglutinin (HA)–tagged Ikaros was readily detectable using anti-HA antibodies and localized to the nucleus (data not shown). As seen in Jurkat cells, synthesis of IL-2 in stimulated Th1 cells was markedly reduced in cells transduced with an Ikaros-expressing retroviral vector (Figure 5B). ChIP assays were also performed in infected cells using an anti-HA antibody. Ikaros binding to the il2 promoter was detected in RV-Ikaros-IRES-GFP–transduced cells compared with control cells (ie, RV-IRES-GFP–infected cells; Figure 5C). ChIP experiments using antibodies against acetylated H4 in Ikaros-expressing Th1 cells were performed to determine whether Ikaros binding could induce histone deacetylation on the il2 promoter. Cells expressing HA-Ikaros showed a profound loss of H4 acetylation on the il2 promoter compared with control cells (Figure 5D). The specificity of this effect was confirmed using primers for the CD3ϵ promoter. Ikaros expression did not affect the H4 acetylation of the CD3ϵ promoter (Figure 5D). To confirm that repression of IL-2 expression by Ikaros was in fact mediated through its capacity to recruit HDACs, the effect of inhibiting HDAC activity with TSA was determined in Th1 cells infected with a retrovirus directing the expression of Ikaros. TSA treatment prevented Ikaros-mediated inhibition of IL-2 expression and restored its production to control levels (Figure 5E).

In order to determine the contribution of Ikaros-mediated il2 repression to T-cell anergy, we analyzed the effect on anergy of blocking Ikaros expression in Th1 cells. To avoid problems resulting from abnormal T-cell development in Ikaros-null mice,¹⁹  wild-type mouse Th1 cells were transfected with a plasmid expressing an siRNA specific for Ikaros mRNA. siRNA-transfected cells were treated with ionomycin to induce unresponsiveness. Reduction of Ikaros expression was determined by Western blot and varied from 40% to 60% (Figure 6A). The defect in IL-2 production in anergic cells was 3-fold smaller in siRNA-transfected cells than the one seen in control transfected cells (P <.01 when comparing means of fold reduction values with a t test for paired samples; Figure 6A). Cells were also retrovirally transduced with a virus expressing Ikaros 6 (Ik6), an isoform of Ikaros that lacks DNA binding activity but is able to dimerize with other Ikaros proteins and therefore block their activity. While anergic control cells underwent deacetylation of the il2 locus, in Ik6-expressing cells, defective induction of anergy correlated with deficient H4 deacetylation. This effect was il2 promoter specific, as no differences were detected when control primers for the CD3ϵ promoter were used (Figure 6B).

Functional inactivation of self-reactive T cells is necessary to ensure proper maintenance of immune tolerance. The molecular mechanisms responsible for the establishment and maintenance of T-cell anergy are only recently beginning to be elucidated. In T cells, clonal anergy may be induced by partial or suboptimal stimulation. Anergizing stimuli activate calcium/Cn/NFAT signals that lead to the induction of a specific set of genes.⁷  Evidence supports that proteins encoded by some of those genes are responsible for the inhibition of T-cell function during anergy. In anergic T cells, signaling pathways activated by TCR engagement may be affected at different levels, leading to inhibition of cell activation, clonal proliferation, or cytokine expression.^26-30  Our results show that an active mechanism of transcriptional repression is engaged in anergic Th1 cells to inhibit IL-2 expression. Anergic T cells induce Ikaros expression that targets the il2 promoter, where it binds and recruits HDACs, inducing epigenetic changes that result in inhibition of il2 expression.

Changes in chromatin structure have been shown to regulate il2 expression in activated T cells. The distal region of the il2 promoter has a partially open conformation in naive T cells, which are already able to synthesize IL-2 upon stimulation.²¹  Increased susceptibility to nucleases of the il2 promoter locus follows T-cell stimulation, which correlates with increased ability to activate il2 transcription.^21,31  Consistent with these data, our results show that in Th1 cells, which have already undergone primary TCR engagement, core H4 acetylation on the il2 promoter is enhanced compared with naive cells. This may ensure a permissive il2 locus able to sustain increased expression following secondary stimulation. Similar regulation of locus accessibility has been described for other cytokines, like IFNγ and IL-4, during T-helper differentiation.^32,33  In both cases, changes in histone acetylation of those genes ultimately determine whether they become active or silent in Th1 or Th2 subpopulations.³³  Partial stimulation of the TCR in the absence of costimulation should not induce further histone acetylation and chromatin remodeling of the il2 locus, as signals activated by CD28 engagement seem to be required for this process.³⁴  We show that epigenetic changes are also established in T cells in response to anergizing stimuli. Active histone deacetylation of the il2 promoter occurs in Th1 cells following partial stimulation, causing stable epigenetic changes in the il2 locus. As in other differentiation processes, epigenetic changes may, thus, contribute to induce a permanent inactive status on the il2 locus and to make inhibition of the il2 gene expression stable in anergic T cells.

We had previously shown that calcium signals activated during anergy induction are responsible for the up-regulation of Ikaros expression in a Cn/NFAT-dependent manner.⁷  Ikaros has been identified as a key regulator of lymphoid lineage development.^17-19  During T-cell development, Ikaros contributes to gene silencing by facilitating the assembly of silent chromatin at specific loci.^18,35,36  Several mechanisms have been proposed for Ikaros' transcriptional repressor activity,^20,37,38  including the recruitment of chromatin remodeling complexes containing HDACs.^20,35,39  The ability of Ikaros to recruit HDACs plays a crucial role in its function as a repressor of gene expression.²⁰  In lymphoid cells, Ikaros associates with NurD and mSIN3 complexes, which contain HDACs, contributing to the control of gene expression during lymphopoiesis.³⁵  Our results indicate that Ikaros may also control mature T-cell function through similar mechanisms. In anergic T cells, Ikaros binds to the il2 promoter and recruits HDAC activity (Figures 4–5). Activity of HDACs seems to be necessary for Ikaros-mediated il2 transcriptional repression because inhibition of HDACs with TSA partially blocks suppression of IL-2 production (Figure 3). Activation of il2 transcription has also been found to be associated with active demethylation of the il2 promoter.⁴⁰  In activated T cells, Ikaros can colocalize with DNA methyltransferases during the S phase of the cell cycle,⁴¹  therefore we cannot rule out the possibility that additional mechanisms might also be involved in silencing il2 gene expression in anergic T cells.

Several mechanisms have been proposed to explain the Ras/mitogen-activated protein kinase (MAPK) pathways block that causes defects in JNK and ERK activation in anergic T cells.^22,42  Hyperactivated Rap1 may compete with Ras for Raf1 binding, preventing activation of MAPKs.⁴³  The involvement of several E3 ubiquitin ligases, including Itch, Nedd4, Grail, and Cbl-b, has also been proposed to explain TCR signaling blockade in anergic T cells. These proteins would add ubiquitin tags to selected proteins involved in TCR signaling, targeting them for degradation, recycling, or inactivation.^11-14  Recently, inactivation of dyacylglycerol-mediated signaling through activation of dyacylglycerol kinase has also been shown to contribute to Ras activity block in anergic T cells.^28,29  All of these mechanism would result in severe signaling defects ranging from altered immune synapse formation to protein lipase C γ1 (PLC-γ1) degradation and Ras inactivation. Stimulation of anergic T cells with PMA and ionomycin has been shown to restore IL-2 production and proliferation. Direct activation of PKC by PMA should overcome proximal blocks. Nevertheless, restoration of T-cell activation in anergic T cells with PMA and ionomycin is only partial.^22,23  It is clear, thus, that other mechanisms, likely downstream of PKC activation, must also contribute to the unresponsive state of anergic T cells.²³  Our results clearly reveal the existence of such a mechanism. Ikaros would induce changes in histone acetylation at the il2 locus. Binding of Ikaros could account, at least in part, for the disparate effects that anergy has on the expression of different cytokines. While IL-2 expression is completely blocked, the expression of IL-4 or IFNγ is only slightly affected²⁷  and IL-10 expression may even be enhanced.⁷  The imposition of a new stable pattern of histone acetylation in the il2 promoter would be responsible for its silencing and, in cooperation with other mechanisms, completely inhibit IL-2 synthesis.

Preventing IL-2 expression is critical to maintain anergy, as signals transmitted by the IL-2 receptor are able to reverse anergy in T cells.^3,44  Binding of CREB/CREM complexes to the −180 AP-1 site of the il2 promoter and il2 repression by Smad3 binding to negative regulatory elements on the il2 promoter have also been proposed to regulate IL-2 expression in anergic T cells, supporting the existence of transcriptional mechanisms that can inhibit il2 expression in anergic T cells.^45,46  Binding of these factors might also be facilitated by changes in histone acetylation in this locus.

Regulation of chromatin accessibility to the il2 promoter may not be restricted only to anergic Th1 cells. It has recently been reported that in CD4⁺CD25⁺ regulatory T cells, restoring defective JNK activation with a constitutively active kinase did not overcome deficient IL-2 expression. Marked differences in chromatin accessibility at the il2 promoter were detected, though, between CD4⁺CD25⁻ and CD4⁺CD25⁺ T cells, which might be able to account for the lack of il2 transcription in regulatory T cells.⁴⁷  It would also be interesting to determine whether other members of the Ikaros family might play similar roles in controlling self-reactivity in T cells or other lymphocytes. We have not detected up-regulation of any other Ikaros family member in anergic Th1 cells, but Aiolos-deficient mice develop an autoimmune phenotype characterized by hyperactivated B cells that spontaneously form germinal centers, increased antinuclear antibody production, and development of a systemic lupus–like syndrome.^48,49 

The importance of keeping self-reactive T cells inactivated to prevent autoimmune reactions may underscore the existence of multiple complementary mechanisms designed to inhibit self-reactive T-cell activation. We have shown that Ikaros induces histone deacetylation of the il2 promoter and inhibits il2 expression in anergic T cells. We propose, thus, that epigenetic modifications can play a crucial role in the establishment and maintenance of T-cell anergy.

Contribution: F.M. and S.B. designed research; S.B., M.D., N.S.-N., and I.P. performed research; S.B. and M.P. contributed new reagents/analytic tools; S.B. and F.M. analyzed data; and F.M. wrote the paper.

Conflict-of-interest: The authors declare no competing financial interests.

Correspondence: Fernando Macian, Albert Einstein College of Medicine, Department of Pathology, 1300 Morris Park Avenue, Bronx, NY 10461; e-mail: fmacianj@aecom.yu.edu.

This work was supported by National Institutes of Health training grants GM007491 (N.S.-N.) and GM007288 (M.P.), the Irene Diamond Fund, and a National Institutes of Health grant AI059738 (F.M.).

We thank members of our laboratory for helpful discussions and A. Jordan and A. M. Cuervo for critical reading of the manuscript. We would like to thank K. Georgopoulos, D. J. McKean, G. P. Nolan, and S. T. Smale for the generous gift of reagents used in this work. Monoclonal antibody against murine IL-4 was obtained form the Biological Resources Branch preclinical repository.