T cell factor 1 (TCF-1) is expressed in both developing and mature T cells and has been shown to restrain mature T cell–mediated Th17 responses by inhibiting IL-17 expression. However, it is not clear when TCF-1 is required in vivo to restrain the magnitude of peripheral Th17 responses and what the molecular mechanisms responsible for TCF-1–regulated IL-17 gene expression are. In this study, we showed that conditional deletion of TCF-1 at the early but not later CD4+CD8+ double-positive stage in mice enhanced Th17 differentiation and aggravated experimental autoimmune encephalomyelitis, which correlates with abnormally high IL-17 expression. Expression of TCF-1 in TCF-1–deficient thymocytes but not TCF-1–deficient Th17 cells inhibited IL-17 expression. TCF-1 binds to IL-17 promoter regions, and deletion of two TCF-1 binding sites relieves TCF-1–mediated inhibition of IL-17 promoter activity. Lastly, wild-type TCF-1, but not a TCF-1 mutant that has no intrinsic histone deacetylase activity, was able to inhibit IL-17 expression in TCF-1 deficient mouse thymocytes. Thus, our study demonstrates the requirement of TCF-1 in vivo at stages earlier than double-positive cells to restrain peripheral Th17 immunity by directly binding and inhibiting IL-17 promoter in its intrinsic histone deacetylase–dependent manner.
T cell development in thymus is a process to arm T cells with the capacity to mediate appropriate immune responses in peripheral tissues. Lymphoid progenitors that developed from hematopoietic stem cells in the bone marrow migrate into the thymus to complete sequential maturation stages, including CD4−CD8− double-negative (DN), CD4+CD8+ double-positive (DP), and CD4+ or CD8+ single-positive stages (1–4). Mature single-positive T cells then migrate to the peripheral lymphoid organs to participate adaptive immune responses against pathogens.
When the single CD4+ positive T cells migrate out of thymus, they are naive and are not competent to mediate immune responses. To become effector T cells, they must undergo an activation and differentiation process. This process is initiated upon encountering Ags and eventually differentiates naive T cells into Th cells that include Th1, Th2, Th17, and regulatory T cells (Treg). Th17 cells secrete IL-17 and participate in protective immunity against pathogens, whereas inappropriately exaggerated Th17 responses contribute to pathological immune responses involved in the autoimmunity, such as psoriasis and multiple sclerosis (5–9).
Besides shaping a T cell repertoire that reacts to foreign but not self-antigens, the thymocyte developmental process also controls the magnitude of T cell responses in the periphery. For example, T cell factor 1 (TCF-1), a transcription factor enriched in hematopoietic cell compartments, regulates T cell development in thymus (10–12). Our previous studies have shown that germline deletion of TCF-1 resulted in increased IL-17 expression both in thymus and peripheral T cells and hence led to enhanced Th17 differentiation and more severe experimental autoimmune encephalomyelitis (EAE) (13, 14), indicating the negative role of TCF-1 in the regulation of Th17 immunity. We have some in vitro evidence that TCF-1 loses the ability to regulate IL-17 gene expression in mature T cells. However, because it was a germline deletion, it is not clear when TCF-1 is required in vivo to limit IL-17 gene expression and thus control the scale of Th17 responses in the periphery. By using conditional deletion of TCF-1 at different developmental stages, we demonstrated that CD4-Cre–mediated deletion of TCF-1 at the CD4+CD8+ DP stage did not significantly affect thymic T cell development, peripheral Th17 differentiation, and EAE. In contrast, Vav1-Cre–mediated deletion of TCF-1 at earlier hematopoietic stages disrupts thymic T cell development and potentiates Th17 differentiation and development of EAE. Moreover, expression of TCF-1 in TCF-1−/− thymocytes but not TCF-1−/− Th17 cells was able to downregulate IL-17 expression. We also found that TCF-1–mediated inhibition of IL-17 expression depends on its intrinsic histone deacetylase (HDAC) activity (15). We mapped the TCF-1 binding regions on the IL-17 promoter, and deletion of the DNA fragments containing the TCF-1 binding sites prevented TCF-1 from inhibiting IL-17 promoter. Therefore, we demonstrated the stage-specific requirement of TCF-1 in vivo during early development to control the scale of the peripheral Th17 immune responses via inhibiting IL-17 expression through the intrinsic HDAC activity of TCF-1.
Materials and Methods
TCF-1fl/fl mice were described previously (16) and obtained from Dr. H.-H. Xue (University of Iowa, Iowa City, IA). Rag1−/−, transgenic CD4-Cre and Vav1-Cre mice were purchased from The Jackson Laboratory. TCF-1fl/fl/CD4-Cre and TCF-1fl/fl/Vav1-Cre were generated by crossing TCF-1fl/fl to CD4-Cre and Vav1-Cre, respectively. For all experiments, mice were 6–10 wk old. All mice were bred and maintained in the specific pathogen-free conditions, and experiments were performed in accordance with a protocol approved by the Institutional Animal Care and Use Committee at the Beckman Research Institute of City of Hope (IACUC 07023).
T cell isolation and in vitro differentiation
Mouse naive CD4+ T cells were isolated from spleens of 6- to 10-wk-old mice by negative selection using a CD4+ T cell isolation kit (Miltenyi Biotec, Bergisch-Gladbach, Germany). T cells were cultured and differentiated in Iscove’s DMEM (Corning) containing 10% FBS, 50 mM 2-ME, and 100 U/ml penicillin-streptomycin at 37°C with 5% CO2. In brief, 4 × 105 cells/well naive CD4+ T cells were first activated with 0.25 μg/ml anti-CD3 (145-2C11; eBioscience) and 1 μg/ml anti-CD28 (37.51; eBioscience) in goat anti-hamster IgG (0.2 mg/ml; MP Biomedicals, Santa Ana, CA) precoated 24-well plates. Cells were then differentiated in the presence of polarizing cytokine and Ab mixture for 72 h. For Th1, the mixture contained 10 ng/ml recombinant murine (rm) IL-2 (BioLegend), 20 ng/ml rmIL-12 (BioLegend), and 20 μg/ml anti-IL4 (11B11; BioLegend). For Th17, the mixture contained 2 ng/ml rmTGF-β1 (eBioscience) and 25 ng/ml rmIL-6 (eBioscience). For Treg, the mixture contained 5 ng/ml rmTGF-β, 10 μg/ml anti-IFN-γ (XMG1.2; BioLegend), and 10 μg/ml anti–IL-4 (11B11; BioLegend).
Flow cytometry and cell sorting
Cultured T cells were stimulated with 50 ng/ml PMA (Sigma-Aldrich, MO) and 1 μg/ml ionomycin (Sigma-Aldrich) in the presence of brefeldin A for 4 h before staining. Both cultured T cells and freshly prepared thymocytes were first stained with fluorescence-conjugated Abs to cell surface markers and then fixed/permeabilized either with the BD Cytofix/Cytoperm kit (BD Biosciences) or Transcription Factor Staining Buffer Set (BD Pharmingen). Cells were then intracellularly strained with fluorescence-conjugated Abs to cytokines or transcription factors. Abs include FITC-CD4 (GK1.5; BioLegend), PE-CD8 (53-6.7; BioLegend), PE-Cy7-CD8 (53-6.7; eBioscience), APC-CD8 (53-6.7; BD Biosciences), PE-Thy1.2 (30-H12; BD Biosciences), PE-TCRβ (H57-597; BD Biosciences), APC-IL-17A (eBio17B7; eBioscience), APC-IFN-γ (XMG1.2; eBioscience), APC-FOXP3 (FJK-16s; eBioscience), and PE-RORγt (Q31-378; BD Biosciences). Flow cytometry was performed by using a FACSCanto II (BD Biosciences) and analyzed with FlowJo software (Tree Star). DN thymocytes (Thy1.2+CD4−CD8−) were electronically sorted by using FACSAria II (BD Biosciences).
Quantitative real-time PCR
Total RNA was prepared using the RNeasy Isolation Kit according to the manufacturer’s instructions (Qiagen). cDNA was synthesized using the Tetro cDNA Synthesis Kit (BIO-65043; Bioline). Quantitative real-time PCR (qRT-PCR) was performed using SsoFast EvaGreen Supermix (Bio-Rad) in a CFX96 Real-Time PCR Detection System (Bio-Rad). Primers are as follows: IL-17A–forward: 5′-TTTAACTCCCTTGGCGCAAAA-3′, IL-17A–reverse: 5′-CTTTCCCTCCGCATTGACAC-3′; IL17F-forward: 5′-TGCTACTGTTGATGTTGGGAC-3′, IL17F-reverse: 5′-AATGCCCTGGTTTTGGTTGAA-3′; CCL20-forward: 5′-GCCTCTCGTACATACAGACGC-3′, CCL20-reverse: 5′-CCAGTTCTGCTTTGGATCAGC-3′; and CCR6-forward: 5′-CCTGGGCAACATTATGGTGGT-3′, CCR6-reverse: 5′-CAGAACGGTAGGGTGAGGACA-3′.
Retroviral packaging and transduction
MSCV-IRES-EGFP–based retroviral expression vectors encoding TCF-1 or empty vectors were transfected into Plat-E packaging cells by Lipofectamine 2000 (Invitrogen Life Technologies) mediated transfection. After 48 h, viral supernatants were collected and stored at −80°C until use. For transduction, naive CD4+ T cells were first activated with 0.25 μg/ml anti-CD3 (145-2C11; eBioscience) and 1 μg/ml anti-CD28 (37.51; eBioscience) in goat anti-hamster IgG precoated 24-well plates for 24 h followed by spin-infection with viral supernatants (2500 rpm, 30°C for 2 h) in the presence of 8 μg/ml polybrene (Sigma-Aldrich). After spin-infection, indicated cytokines described above were added to the culture media to induce Th17 differentiation.
Coculture and transduction of thymocytes on OP9-DL1 stromal cells
Thymocytes were allowed to differentiate on OP9-DL1 cells (gift from Dr. T.P. Bender, University of Virginia Health System, Charlottesville, VA). Briefly, 5 × 105 cells/well electronically sorted Thy1.2+CD4−CD8− thymocytes were cultured overnight on an 80% confluent OP9-DL1 monolayer in the flat-bottom 24-well plates in αMEM (Stemcell Technologies) supplemented with 20% FBS, 100 U/ml penicillin-streptomycin, 2 mM l-glutamine (Invitrogen Life Technologies), and 5 ng/ml rmIL-7 (PeproTech). Cocultures were then spin-infected (2500 rpm, 30°C for 2 h) with retroviral supernatants in the presence of 5 μg/ml polybrene. Seventy-two hours posttransduction, cocultures were harvested for flow cytometry.
Induction and assessment of EAE
Active EAE was induced according to the manufacturer’s instructions (Hooke Laboratories, Lawrence, MA). Briefly, mice were immunized with 200 μl of myelin oligodendrocyte glycoprotein 35–55 (MOG35–55) peptide emulsion subcutaneously. On days 0 and 1 after immunization, mice were injected i.p. with 200 ng of Bordetella pertussis toxin (Hooke Laboratories). For Th17- or Th1-induced passive EAE, donor mice were immunized with MOG35–55 subcutaneously. Ten days later, cells were isolated from spleen and lymph node and cultured with 20 μg/ml MOG35–55 for 3 d under either Th17-polarizing conditions (20 ng/ml rmIL23; R&D) or Th1-polarizing conditions (20 ng/ml rmIL-12, R&D; 2 μg/ml α-IL23p19, eBioscience) (17). Rag1−/− recipient mice were then transferred i.p. with 3.0 × 107 MOG35–55–specific Th17 or Th1 cells. The severity of EAE was monitored and evaluated on a scale from 0 to 5 according to Hooke Laboratories’ guideline. In brief, 0 = no disease, 1 = paralyzed tail, 2 = hind limb weakness, 3 = hind limb paralysis, 4 = hind and fore limb paralysis, and 5 = moribund and death. When a mouse was euthanized because of severe paralysis, a score of 5 was entered for that mouse for the rest of the experiment.
Isolation and analysis of CNS infiltrating cells
Brain and spinal cords were homogenized and single-cell suspensions were prepared using 70-μm cell nylon cell strainers (Fisher Scientific). Cells were collected by centrifugation at 400 relative centrifugal force for 10 min. Cells were then resuspended in 4ml of 30% Percoll (GE Healthcare) and centrifuged onto 5 ml of 70% cushion for 20 min at 2000 rpm. Lymphocytes at 30–70% interface were collected and subjected to cell surface staining with fluorescence-conjugated Abs: FITC-CD45 (30-F11; eBioscience), APC-CD3 (145-2C11; BD Pharmingen), APC-Cy7-Ly6C (HK1.4; eBioscience), PE-Cy7-Ly6G (1A8; BioLegend), and PE-CD11b (M1/70; eBioscience). For intracellular cytokine staining, the CNS infiltrating cells were stimulated with 50 ng/ml PMA (Sigma-Aldrich) and 1 μg/ml ionomycin (Sigma-Aldrich) in the presence of brefeldin A for 4 h and were then stained with PE-Cy7-CD4 followed by intracellular staining of FITC-IFN-γ (XMG1.2; BioLegend), PE-GM-CSF (MP1-22E9; BioLegend), and APC-IL-17A (eBio17B7; eBioscience).
Chromatin immunoprecipitation on qRT-PCR
Chromatin immunoprecipitation (ChIP) on qRT-PCR was performed with the ChIP-IT High Sensitivity Kit (53040; Active Motif). Briefly, 1.5 × 107 cells per ChIP were obtained from thymus or differentiated Th17 cells. Cells were fixed and sheared as described in the ChIP-IT High Sensitivity Kit manual. ChIP reactions were then performed on 20 μg of the prepared chromatin using 4 μg of anti–TCF-1 Ab (C63D9; Cell Signaling) or anti–acetyl-histone H3 (06-599; Millipore). Four micrograms of IgG was used as negative control. Input DNA and DNA pull-down after immunoprecipitation were analyzed by qRT-PCR as described above. Primers that cover ∼2000 bp of the IL-17A promoter region were designed as follows: P1-forward: 5′-AAACCCTATGCAGTTGGTACA-3′, P1-reverse: 5′-CATCTCTCCAGCTCCATGGA-3′; P2-forward: 5′-AGCCCATAAAGAAGCCAATGT-3′, P2-reverse: 5′-TCCTGGCTTTCTACGTGTCA-3′; P3-forward: 5′-GGGTGAAAGAGGACATTGCC-3′, P3-reverse: 5′-TGCCAGGTCTTTTCCCATTC-3′; P4-forward: 5′-TGACACGTAGAAAGCCAGGA-3′, P4-reverse: 5′-GATTGCCAGGTCTTTTCCCA-3′; P5-forward: 5′-GGCAATCAGAGGTGTGTGTG-3′, P5-reverse: 5′-CTCAGAAGTTGCAGCACCTC-3′; P6-forward: 5′-CTATCGGTCCACCTCATGCT-3′, P6-reverse: 5′-GGAGATGAGGGATGAGAAGGG-3′; P7-forward: 5′-GTTAGTAGTCTCCACCCGGC-3′, P7-reverse: 5′-TGAGGTTCGGTATCAAGCCT-3′; P8-forward: 5′-GAGTGGGTTTCTTTGGGCAA-3′, P8-reverse: 5′-AGCATGACTTCTTGGGAGCT-3′; and P9-forward: 5′-AGCTCCCAAGAAGTCATGCT-3′, P9-reverse: 5′-TACGTCAAGAGTGGGTTGGG-3′.
Vav1-Cre– but not CD4-Cre–mediated deletion of TCF-1 potentiates Th17 differentiation and EAE
Previous studies of TCF-1 function mostly depend on using a strain of germline knockout mice that cannot determine the tissue- and time-specific function of TCF-1 (10). To determine the critical stages in which TCF-1 is required to restrain peripheral Th17 responses, we crossed conditional TCF-1 (TCF-1fl/fl) mice (16) to CD4-Cre and Vav1-Cre mice, respectively. CD4-Cre induces gene deletion at the CD4+CD8+ DP stage (18), whereas Vav1-Cre induces deletion at earlier hematopoietic stages, including the CD4−CD8− DN stage (19). We performed PCR analysis of the TCF-1 locus to detect the deletion of the TCF-1 gene (Fig. 1A). As expected, there was no deletion at all in the absence of Cre, as shown in TCF-1fl/fl mice. However, TCF-1 deletion was detected in CD4+CD8+ thymocytes and peripheral T cells but not in CD4−CD8− thymocytes of the TCF-1fl/fl/CD4-Cre mice. In contrast, TCF-1fl/fl/Vav1-Cre mice deleted TCF-1 in CD4−CD8−, CD4+CD8+, and peripheral T cells, confirming that Vav1-Cre induced TCF-1 deletion at the stage earlier than CD4−CD8− DN cells. To determine the effects of deletion of TCF-1 at different developmental stages on peripheral Th17 responses, EAE was induced (Fig. 1B). Interestingly, no obvious differences of EAE development in terms of disease onset and severity were observed between TCF-1fl/fl and TCF-1fl/fl/CD4-Cre mice, with a comparable mean peak disease score of ∼3. However, the clinical score of TCF-1fl/fl/Vav1-Cre mice reached over 3.5, indicating more severe EAE than that observed in TCF-1fl/fl/CD4-Cre mice. These results demonstrated that deletion of TCF-1 at earlier stages, but not at later DP stages, potentiated Th17 immunity responsible for EAE. Moreover, TCF-1fl/fl/Vav1-Cre mice had a significantly reduced number of peripheral T cells compared with TCF-1fl/fl and TCF-1fl/fl/CD4-Cre mice (Fig. 1C), suggesting fewer T cells, but with higher potency in the induction of EAE when TCF-1 was deleted by Vav1-Cre.
Because Th17 cells contribute to EAE development (20), we wondered how stage-specific deletion of TCF-1 affects Th17 differentiation. Even under Th0 conditions, we have already observed an increased percentage of IL-17+ cells in naive CD4+ T cells from TCF-1fl/fl/Vav1-Cre mice, whereas IL-17 expression was hardly detectable in TCF-1fl/fl and TCF-1fl/fl/CD4-Cre T cells. Under Th17 polarization conditions, T cells from TCF-1fl/fl and TCF-1fl/fl/CD4-Cre mice were equivalent in the generation of IL-17+ cells, whereas ∼100% more IL-17+ cells were developed from TCF-1fl/fl/Vav1-Cre T cells (Fig. 1D). Increased IL-17+ TCF-1fl/fl/Vav1-Cre T cells correlated with the significant increased expression of Th17 signature genes, including IL-17A, IL-17F, CCR6, and CCL20 (Fig. 1E). In contrast, naive CD4+ T from TCF-1fl/fl/Vav1-Cre mice had no defects in Th1 and Treg differentiation (Supplemental Fig. 1), suggesting the selective role of TCF-1 in Th17 differentiation. These results demonstrate that TCF-1 is required at early but not later CD4+CD8+ stages during T cell development to restrain peripheral Th17 immunity by selectively controlling the potential of Th17 differentiation.
Deletion of TCF-1 by Vav1-Cre potentiates Th17- but not Th1-mediated EAE
TCF-1 deficiency skews Th2 fate to Th1 fate (21). Besides Th17 cells, Th1 cells can also contribute to EAE induction, but these two forms of EAE display distinct lesion sites and immune profiles in CNS (22, 23), indicating different mechanisms are involved in the development of Th17- and Th1-mediated EAE. Although our in vitro experiment did not show significant difference in Th1 differentiation between TCF-1fl/fl/Vav1-Cre mice and TCF-1fl/fl and TCF-1fl/fl/CD4-Cre control (Supplemental Fig. 1), it is not clear whether deletion of TCF-1 affects Th1 function in term of induction of EAE in vivo. We thus compared the function of differentiated Th1 and Th17 cells by means of induction of passive EAE. Interestingly, Rag1−/− recipient mice (which have a congenital deficiency in mature B and T cells) reconstituted with cells from spleen and lymph node of TCF-1fl/fl, TCF-1fl/fl/CD4-Cre, or TCF-1fl/fl/Vav1-Cre mice stimulated under Th1 priming conditions developed equivalent EAE (Fig. 2A, 2B). However, cells from TCF-1fl/fl/Vav1-Cre mice stimulated under Th17 priming conditions induced more severe EAE in Rag1−/− recipient mice (Fig. 2C), which was also indicated by a significantly increased absolute number of CNS-infiltrating lymphocytes, including T cells, neutrophils, and monocytes (Fig. 2D, 2E, gate strategy for different cells shown in Supplemental Fig. 2). In addition, intracellular staining of CNS-infiltrating cells showed a greatly increased percentage of CD4+IL-17+IFN-γ+ and CD4+IL-17+GM-CSF+ cells, which are believed to be pathogenic for EAE (5–7), in the CNS of Rag1−/− mice adoptively transferred with Th17-priming conditions–treated cells from TCF-1fl/fl/Vav1-Cre mice (Fig. 2F, 2G). Taken together, these findings supported that the exaggerated EAE resulted from deletion of TCF-1 at an early T cell developmental stage was because of enhanced Th17 but not Th1 responses.
Vav1-Cre– but not CD4-Cre–induced deletion of TCF-1 leads to increased IL-17 expression in thymus
Because we observed increased IL-17 expression in mature CD4+ T cells from TCF-1fl/fl/Vav1-Cre mice (Fig. 1D), we next wondered whether these increased IL-17–producing cells originated from thymus. Vav1-Cre– but not CD4-Cre–induced deletion of TCF-1 affected thymocyte development indicated by thymic weight (Fig. 3A, left panel) and cellularity (Fig. 3A, right panel) as well as developmental stages defined by CD4 and CD8 surface markers (Fig. 3B), suggesting that TCF-1 plays a major role at rather earlier developmental stages to ensure normal thymocyte development. Intracellular IL-17 staining indicated that more mature CD4+ TCRhi thymocytes from TCF-1fl/fl/Vav1-Cre mice produce IL-17 (3.10%) than those from TCF-1fl/fl (0.03%) and TCF-1fl/fl/CD4-Cre (0.2%) mice (Fig. 3B). This was consistent with our observation that peripheral CD4+ cells from TCF-1fl/fl/Vav1-Cre mice generated more IL-17–producing cells, even in Th0-neutral conditions (Fig. 1D). When gated on an earlier CD4−CD8− TCR− stage, we still found that more CD4−CD8− thymocytes from TCF-1fl/fl/Vav1-Cre mice produced IL-17 (4.51%) compared with those from TCF-1fl/fl (0.42%) and TCF-1fl/fl/CD4-Cre (0.48%) mice (Fig. 3B). Therefore, Vav1-Cre–induced deletion of TCF-1 resulted in heightened IL-17 expression at as early as the CD4−CD8− DN stage, and such heightened IL-17 expression was maintained at the mature CD4+ T cell stage, even after migrating out of thymus to the periphery. Because RORγt is required to stimulate IL-17 expression (24), RORγt expression was assessed at different thymocyte developmental stages in all three groups (Fig. 3C). Consistent with previous observations (25), RORγt was upregulated at the CD4+CD8+ DP stage but downregulated at the mature CD4+ stage in control TCF-1fl/fl mice. RORγt expression levels overlapped well between TCF-1fl/fl and TCF-1fl/fl/CD4-Cre thymocytes at CD4−CD8−, CD4+CD8+, and CD4+ stages. In contrast, a portion of CD4−CD8− DN and CD4+ thymocytes expressed higher levels of RORγt in TCF-1fl/fl/Vav1-Cre mice (Fig. 3C, arrows), which correlated with the observed increased IL-17 expression. Taken together, these data indicate that TCF-1 is required at early but not late developmental stages to inhibit IL-17 expression during T cell development.
Forced expression of TCF-1 in CD4−CD8− thymocytes of TCF-1fl/fl/Vav1-Cre mice inhibits IL-17 expression
We next determined whether forced expression of TCF-1 could inhibit IL-17 expression in developing and mature T cells from TCF-1fl/fl/Vav1-Cre mice. Just like in mature T cells with germline deletion of TCF-1 (13), expression of TCF-1 by retrovirus-mediated gene transduction in mature T cells from TCF-1fl/fl/Vav1-Cre mice was unable to reduce IL-17 expression to wild-type (WT) levels under Th17 priming conditions (Fig. 4A). Next, we determined the effects of TCF-1 on thymocytes cocultured with OP9-DL1 stroma cells (26). TCF-1fl/fl CD4−CD8− DN thymocytes could develop to CD4+CD8+ DP thymocytes with or without expression of exogenous TCF-1 from retrovirus (Fig. 4B). TCF-1fl/fl/CD4-Cre CD4−CD8− DN thymocytes were able to develop to CD4+CD8+ DP thymocytes, however, with reduced efficiency indicated by a reduced percentage of CD4+CD8+ cells. Exogenous TCF-1 could restore the percentage of CD4+CD8+ cells in TCF-1fl/fl/CD4-Cre thymocytes to WT level (Fig. 4B), indicating the intact function of exogenous TCF-1. In contrast to TCF-1fl/f and TCF-1fl/fl/CD4-Cre thymocytes, the majority of TCF-1fl/fl/Vav1-Cre CD4−CD8− DN thymocytes stayed at the CD4−CD8− DN stage, even in the presence of exogenous TCF-1 (Fig. 4B). We next determined the effects of exogenous TCF-1 on IL-17 expression. IL-17 was hardly detectable in thymocytes (including DN and developed DP) from TCF-1fl/fl and TCF-1fl/fl/CD4-Cre mice with or without exogenous TCF-1 (Fig. 4C), whereas CD4−CD8− DN thymocytes from TCF-1fl/fl/Vav1-Cre mice had abundant IL-17 expression (Fig. 4C), consistent with what we observed in vivo (Fig. 3B). Indeed, exogenous TCF-1 significantly suppressed IL-17 expression in TCF-1fl/fl/Vav1-Cre thymocytes (Fig. 4C). Taken together, these data further demonstrated that TCF-1 is required to suppress IL-17 expression at early T cell developmental stages.
TCF-1 binds to IL-17 promoter and inhibits IL-17 expression through its intrinsic HDAC activity
Histone deacetylation is an evolutionarily conserved epigenetic modification mechanism responsible for gene repression (27). Recent studies have shown that TCF-1 had intrinsic HDAC activity (15). Considering the dual capacity of DNA binding and intrinsic HDAC activity in TCF-1, we wondered whether TCF-1 binds to IL-17 locus and silences IL-17 expression via its intrinsic HDAC activity. To answer this question, we first determined whether TCF-1 binds to IL-17 promoter by ChIP assays with nine primer pairs covering the IL-17A promoter regions (P1 to P9) ∼2000 bp upstream of the transcriptional starting site (Fig. 5A). The enrichment of TCF-1 binding to IL-17 promoter was observed in the region from P4 to P8 in TCF-1fl/fl thymus versus differentiated Th17 cells (Fig. 5B). To determine which TCF-1 binding regions are functional, each individual region from P4 to P8 was deleted to assess the effects on IL-17 reporter activity. Consistent with previous results (28, 29), RORγt stimulated IL-17 reporter. Whereas TCF-1 inhibited RORγt-dependent IL-17 reporter, deletion of the P4 or P7 regions, but not the P5, P6, or P8 regions, of IL-17 promoter significantly relieved TCF-1–mediated inhibition of IL-17 reporter (Fig. 5C), suggesting that P4 and P7 are two functional TCF-1 binding sites responsible for the inhibition of IL-17 promoter activity. To determine whether TCF-1 is required for epigenetic modification critical for IL-17 expression, we compared histone H3 acetylation levels in the regions from P4 to P8. Because of the deletion of TCF-1 in TCF-1fl/fl/Vav1-Cre thymus, P5 and P8 had higher H3 acetylation levels (Fig. 5D), indicating that these two regions underwent TCF-1–dependent histone deacetylation in the WT thymus. These data indicate the possibility that TCF-1 mediates histone deacetylation in the regions around these binding sites. We further observed that, in contrast to WT TCF-1, TCF-1-5aa mutant, which diminished HDAC activity (15), was unable to inhibit IL-17 expression in DN thymocytes obtained from TCF-1fl/fl/Vav1-Cre mice, which is comparable to virus expression of GFP alone (Fig. 5E). Taken together, these results demonstrated that TCF-1 binds to IL-17 promoter and silences IL-17 expression through its intrinsic HDAC activity.
Previous studies have shown TCF-1 was able to restrain peripheral Th17 responses, and in vitro experiments show that TCF-1 negatively regulates peripheral Th17 responses at T cell developmental stages in thymus rather than at the mature stage of T cells in the periphery (13, 14). However, there was no in vivo evidence to support stage-specific requirement of TCF-1 for limiting peripheral Th17 immunity. Furthermore, the mechanisms responsible for TCF-1–mediated inhibition of Th17 responses remain unknown. In this study, we answered the critical question of when TCF-1 is required to control the peripheral Th17 responses, given that T cell development is a sequential multistep process. We found that deletion of TCF-1 early in the hematopoietic stage, but not in later DP stages, resulted in exaggerated Th17 differentiation and EAE development. In addition, deletion of TCF-1 at early but not later DP stages leads to significant disruption of thymocyte development, demonstrating that TCF-1 plays a major role during early thymocyte development. Mechanically, we show that TCF-1 binds to and inhibits IL-17 promoter to prevent IL-17 expression via its intrinsic HDAC activity.
The IL-17 gene locus must be kept inactive by TCF-1 prior to developing thymocytes entering the DP stage. This is indicated by the fact that deletion of TCF-1 at the DP stage does not cause enhanced IL-17 expression in thymus, which means that the suppressive function of TCF-1 for the IL-17 gene is not required at the DP stage or later. Once TCF-1–mediated inhibition is imprinted on the IL-17 gene locus, which we believe occurred no later than the DN stage, it can last even after T cell migration out of the thymus, resulting in prevention of exaggerated Th17 responses in the periphery. Furthermore, our results highlight the specific requirement of HDAC activity of TCF-1 for suppression of IL-17 gene activity during T cell development. The HDAC family consists of a large number of family members, many of which play indispensable roles in regulating T cell development, including DN to DP transition, cell survival, and TCR signaling in DP thymocytes (30–33). Although several HDAC proteins are expressed in thymus, it seems they are unable to compensate for the HDAC activity of TCF-1, as indicated by abnormal high levels of IL-17 expression (or natural Th17 cells) in thymus of TCF-1fl/fl/Vav1-Cre mice. This phenomenon highlights the specific requirement of HDAC activity of TCF-1 for keeping the IL-17 gene inactive early during T cell development.
Interestingly, if the IL-17 gene locus does not undergo TCF-1–mediated inhibition like TCF-1 deficiency at the early development stage, it will lead to an elevated Th17 differentiation and Th17 immunity. This consequence is irreversible, as reconstitution of TCF-1 in mature T cells was unable to reduce IL-17 expression to normal levels. The mechanisms of how TCF-1 closes the IL-17 gene locus in early developing thymocytes rather than in mature T cells remains largely unknown. Our results suggest that the IL-17 gene locus in mature T cells was relatively inaccessible for TCF-1 compared with that in the early T cell developmental stage. This may result from chromatin remolding during later developmental stages and/or lack of cofactors that present in early thymocytes but not in mature T cells, which could be an interesting topic for future investigation.
Exaggerated Th17 responses are associated with many types of autoimmune diseases (8, 34). To develop effective treatment for autoimmune diseases, many studies have focused on understanding the mechanisms of regulating Th17 differentiation and Th17–Treg balance in the peripheral immune system. Our study suggests that peripheral Th17 responses can be greatly influenced by specific developmental events in the central immune system, thus providing an additional layer of control to prevent autoimmune diseases.
We thank Dr. Hai-Hui Xue for generously sharing TCF-1fl/fl mice. We appreciate the help by City of Hope–supported cores, including animal, genomic, and flow cytometer cores. In addition, we thank Chris Gandhi and Keely Walker for their careful proofing/editing work on the manuscript.
This work was supported by National Institutes of Health Grants R01-AI053147 and R01-AI109644 and institutional pilot funding. Research reported in this publication was also supported by the National Cancer Institute of the National Institutes of Health under Award P30CA33572, which includes work performed in the animal, genomic, flow cytometer, and mass spectrometric cores supported by this grant. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
The online version of this article contains supplemental material.
Abbreviations used in this article:
The authors have no financial conflicts of interest.