Abstract
Histone acetyltransferases (HATs) regulate inducible transcription in multiple cellular processes and during inflammatory and immune response. However, the functions of general control nonrepressed–protein 5 (Gcn5), an evolutionarily conserved HAT from yeast to human, in immune regulation remain unappreciated. In this study, we conditionally deleted Gcn5 (encoded by the Kat2a gene) specifically in T lymphocytes by crossing floxed Gcn5 and Lck-Cre mice, and demonstrated that Gcn5 plays important roles in multiple stages of T cell functions including development, clonal expansion, and differentiation. Loss of Gcn5 functions impaired T cell proliferation, IL-2 production, and Th1/Th17, but not Th2 and regulatory T cell differentiation. Gcn5 is recruited onto the il-2 promoter by interacting with the NFAT in T cells upon TCR stimulation. Interestingly, instead of directly acetylating NFAT, Gcn5 catalyzes histone H3 lysine H9 acetylation to promote IL-2 production. T cell–specific suppression of Gcn5 partially protected mice from myelin oligodendrocyte glycoprotein–induced experimental autoimmune encephalomyelitis, an experimental model for human multiple sclerosis. Our study reveals previously unknown physiological functions for Gcn5 and a molecular mechanism underlying these functions in regulating T cell immunity. Hence Gcn5 may be an important new target for autoimmune disease therapy.
Introduction
Acetylation of either histones or transcription factors is generally associated with transcriptional activation, which plays key roles in regulating T cell development, activation, and differentiation. Acetylation is catalyzed by histone acetyltransferases (HATs), which add acetyl groups to lysine residues in their targets. Several HATs have been identified as transcription coactivators, including CBP, p300, and Tip60, all of which are involved in a variety of biological functions (1). The general control nonrepressed–protein 5 (Gcn5) was initially identified as a member of the HAT superfamily that positively regulates the transcription of amino acid biosynthetic genes in yeast (2–4). Grant et al. (5) defined Gcn5 as the catalytic subunit of a transcriptional coactivation complex SAGA (Spt-Ada-Gcn5-acetyltransferase) that is involved in regulating gene transcription in different tissues and organs. Germline Gcn5 gene deletion in mice leads to embryonic lethality because of increased apoptosis and mesodermal defects (6–8). Therefore, Gcn5 is a critical regulator in a variety of biological, developmental, and pathological functions.
As an HAT, Gcn5 has been shown to regulate gene transcription by catalyzing the acetylation of lysine residues on multiple histones including H2B, H3, and H4 (9–11). In addition to histones, Gcn5 can directly interact with and acetylate transcription factors in gene transcriptional regulation (12–16). Recent in vitro studies suggest that Gcn5 is a critical survival factor during the development and activation of B cells (17, 18), and that Gcn5 regulates CD4+ Th cell differentiation toward IL-9–producing Th9 cells by activating the transcription factor PU.1 (19). However, the in vivo physiological functions of Gcn5 in T cell immunity remain uncharacterized.
In this study, we generated a strain of mice with a T cell–specific Gcn5 gene deletion and discovered that Gcn5 is required for both T cell development and activation through interacting with NFAT. Interestingly, instead of catalyzing NFAT acetylation, Gcn5 is recruited onto the il-2 promoter by NFAT, and it catalyzes the acetylation of lysine residue 9 of histone H3 (H3K9) to regulate IL-2 production during T cell activation. Our study reveals important functions of Gcn5 in T cell immunity in vivo, as well as the underlying molecular mechanisms.
Materials and Methods
Cells, reagents, Abs, and plasmids
HEK293 cells were cultivated in DMEM with 10% of FBS. Inhibitors that suppress calcineurin, cyclosporine A (CsA); JNK1, SP600125; and NF-κB, JHS-23 were purchased from EMD (San Diego, CA). Specific Abs against Gcn5, NFAT1, and HA were from Santa Cruz (Santa Cruz, CA) and against acetylated H3K9 and histone H3 were from Cell Signaling (Cambridge, MA). Fluorescence-conjugated Abs used for cell surface marker analysis and intracellular staining including CD4, CD8, CD25, CD44, IL-2, Foxp3, IFN-γ, IL-4, and IL-17, as well as those for ELISA analysis including IL-2, IL-4, IL-17, IFN-γ, and Abs against each specific isotype of mouse Ig were from eBioscience (San Diego, CA). Gcn5 expression plasmid was purchased from Addgene (Cambridge, MA), and HA-NFAT1 is a gift of Rao’s laboratory (20).
Mice
Gcn5 floxed mice were used as described previously (21, 22). Mice have been backcrossed onto the C57/BL6 genetic background for seven generations. T cell–specific Gcn5-null mice were generated by breeding Gcn5 floxed mice with Lck-Cre transgenic mice as reported previously (23). Gcn5loxp/loxpUB/ESR-CreTG mice were generated by breeding Gcn5 floxed mice with UB/ESR-Cre transgenic mice. All mice used in this study were maintained and used at the Northwestern University mouse facility under pathogen-free conditions according to institutional guidelines and animal study proposals approved by the Institutional Animal Care and Use Committee.
Flow cytometry analysis and ELISA
For the cell surface marker analysis, single-cell suspension was isolated from thymus and spleen of wild-type (WT) and Gcn5 conditional knockout (KO) mice, and stained with fluorescence-conjugated Abs against each specific cell surface marker, including CD3, CD28, CD25, CD44, CD69, CD62L, IL-7Rα, and IL-15Rα (all from eBioscience) as indicated on ice for 30 min, washed, fixed in 1% paraformaldehyde, and analyzed by flow cytometry as indicated (24, 25). For intracellular staining, cells were stimulated with PMA (20 ng/ml) and ionomycin (500 ng/ml) in the presence of brefeldin A (10 μg/ml) for 4 h; the cytokine production was analyzed by intracellular staining following the manufacturer’s protocol followed by flow cytometry analysis. In addition, the levels of cytokine in the culture supernatants were determined by ELISA as described previously (26).
T cell activation and differentiation in vitro and in vivo
For the in vitro T cell activation analysis, naive CD44loCD62Lhi naive CD4+ T cells were sorted from the spleens of WT and KO mice, stained with CFSE, and cultivated with anti-CD3 (1 μg/ml) plus anti-CD28 (1 μg/ml) or under each polarization condition as described previously (27). T cell proliferation was determined by flow cytometry analysis of CFSE dilution or [3H]thymidine incorporation. The production of cytokines including IL-2, IFN-γ, IL-4, and IL-17 was examined by intracellular staining or ELISA.
For the in vitro–induced Gcn5 deletion by tamoxifen, naive CD4 T cells were purified from the spleen of Gcn5loxp/loxpUB/ESR-CreTG mice and then cultivated with 20 ng/ml mouse rIL-7 (eBioscience) with or without (used as WT controls) 20 μg/ml tamoxifen for 7 d. The expression levels of Gcn5 were confirmed by Western blotting, and their proliferation was analyzed.
For the in vivo Ag-specific immune responses, 8- to 10-wk-old WT and Gcn5 conditional KO mice were immunized s.c. with 200 μg of chicken OVA emulsified with 50 μl of CFA at day 1 and boosted with 100 μg of OVA with 50 μl of IFA at day 8. Sera were collected at days 7 and 14 for the analysis of OVA-specific Ig levels. Immunized mice were euthanized at day 14. Total splenocytes were cocultivated with different doses of OVA (0–500 ng/ml) for 3 d. The OVA-specific T cell proliferation was determined by CFSE staining, and their cytokine production was determined by intracellular staining as reported previously (27).
For adoptive transfer, CD45.1 WT and CD45.2 Gcn5-null naive T cells were isolated, mixed at 1:1 ratio, and adoptively transferred into TCRβδ knockout mice (002122; Jackson Laboratory). Seven days after adoptive transfer, recipient mice were euthanized. Single-cell suspension of the solenocytes was prepared, and CD4 and CD8 T cells were analyzed by flow cytometry.
Coimmunoprecipitation and Western blotting analysis of protein expression and interaction
Transient transfection of HEK 293T cells was performed by using Lipofectamine 2000 (Invitrogen). Two days after transfection, cells were collected and lysed in Nonidet P-40 lysis buffer (1% Nonidet P-40, 20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 5 mM EDTA, and freshly added protease inhibitor cocktails). The cell lysates were precleaned and then incubated with Abs (1 μg) for 2 h on ice, followed by the addition of 30 μl of fast-flow protein G-Sepharose beads (GE Healthcare Bioscience, Uppsala, Sweden) overnight at 4°C. Immunoprecipitates were washed four times with NP-40 lysis buffer and boiled in 30 μl of 2× Laemmli buffer. Samples were subjected to 8 or 10% SDS-PAGE analysis and electrotransferred onto nitrocellulose membranes (0.45 μM; Bio-Rad). Membranes were probed with the indicated primary Abs, followed by HRP-conjugated secondary Abs. Membranes were then washed and visualized with an ECL detection system (Bio-Rad, Hercules, CA). When necessary, membranes were stripped by incubation in stripping buffer (Thermo, Rockford, MA), washed, and then reprobed with other Abs as indicated (28–30).
Chromatin immunoprecipitation
Naive CD4+ T cells isolated from WT and Gcn5 conditional KO mice were stimulated with anti-CD3 and anti-CD28 for 16 h. The following inhibitors from Calbiochem (San Diego, CA) were used at each indicated concentration: calcineurin-specific inhibitor CsA, 100 ng/ml; JNK inhibitor SP60025, 10 μg/ml; and NF-κB inhibitor JHS-23, 20 μg/ml. The activated T cells were washed with PBS and cross-linked with 10% formalin, and subjected to chromatin immunoprecipitation (ChIP) using the Chromatin Immunoprecipitation Assay Kit (Millipore) per manufacturer’s instructions. In brief, 2 × 106 cells were lysed with SDS lysis buffer. Cell lysates were sonicated, and 3% of cell lysate was removed and used to determine the total amount of target DNA in input. Remaining cell lysates were diluted in ChIP dilution buffer. Immunoprecipitation was performed with each of the indicated Abs (3 μg) at 4°C overnight. Immune complexes were then mixed with salmon sperm DNA/protein agarose 50% slurry at 4°C for 1 h. After immunoprecipitates were washed sequentially with low-salt buffer, high-salt buffer, LiCl wash buffer, and Tris EDTA, DNA–protein complexes were eluted with elution buffer, and cross-linking was reversed. Genomic DNA was extracted using phenol/chloroform, and ethanol-precipitated DNA was resuspended in Tris EDTA. Quantitative PCR was performed with specific primers as listed in Supplemental Table I.
Experimental autoimmune encephalomyelitis induction and isolation of infiltrated lymphocytes
Experimental autoimmune encephalomyelitis (EAE) was induced and characterized as previously reported (31). Six- to 8-wk-old C57/BL6 mice were immunized with myelin oligodendrocyte glycoprotein (MOG)35–55 peptide (200 μg per mouse emulsified with CFA). Mice were also given pertussis toxin (200 ng per mouse) on days 0 and 2 via tail-vein injection. All mice were weighed and examined daily for clinical symptoms and assigned scores on a scale of 0 to 5 as follows: 0, no overt signs of disease; 1, limp tail; 2, limp tail and partial hind-limb paralysis; 3, complete hind-limb paralysis; 4, complete hind-limb and partial forelimb paralysis; or 5, moribund state or death.
Mice were euthanized during the peak of disease. After extensive transcardial perfusion with PBS, spinal cords and brains were removed from mice (tissues from three mice were pooled for each experiment) and pooled for mononuclear cell isolation. In brief, tissues were incubated with collagenase D (300 μg/ml) and DNase I (20 μg/ml) in HBSS. After 45 min at 37°C, tissues were mechanically dissociated through a 40-μm strainer and washed with PBS. The resultant pellet was fractionated on a discontinuous Percoll gradient. Infiltrating mononuclear cells were harvested from the interface, washed, counted, and cultured for 4 h in RPMI 1640 containing 10% FCS with PMA (10 ng/ml) and ionomycin (500 ng/ml) in the presence of brefeldin A (10 μg/ml) for flow cytometry.
Results
Gcn5 gene deletion impairs T cell activation
To study the physiological functions of Gcn5 in T cell immunity, we conditionally deleted the Gcn5 gene in developing thymocytes in mice by crossing Gcn5-floxed and Lck-Cre mice as reported. Cre recombinase expression under the Lck promoter drives Gcn5 gene deletion at the CD4 CD8 double-negative (DN) stage 2 (DN2) during T cell development (32). Because it has been reported that Cre expression affects T cell functions, Gcn5+/+Lck-Cre+ mice were used as controls. Gcn5-targeted mice were generated by flanking exons 3–19 with two loxP sites (33). Cre recombinase expression leads to a complete absence of Gcn5 protein expression. Indeed, Western blotting analysis confirms a complete loss of Gcn5 protein expression in T cells isolated from the spleen of Gcn5f/fLck-Cre+ mice (Fig. 1A), and the complete Gcn5 deletion was further confirmed by real-time RT-PCR (Fig. 1C). Compared with control littermates, mice with T cell–specific Gcn5 gene deletion have an ∼50% reduction in total cell number in the thymus (Fig. 1D). Although the percentages of cells at CD4 and CD8 double-positive (DP) and single-positive (SP) stages were not altered, their absolute numbers were significantly reduced in the thymus of Gcn5f/fLck-Cre+ mice (Fig. 1D–F). In contrast, only a slight increase in the percentage (Fig. 1E) of CD4−CD8− cells was observed upon Gcn5 gene deletion, indicating that Gcn5 loss most impairs T cell development during the transition stage from DN to DP (Fig. 1F). Further characterization of CD4 and CD8 DN cells by their expression of CD44 and CD25 detected an increase in CD25+CD44− DN stage 3 (DN3) but a reduction in CD25−CD44− DN stage 4 (DN4) cells (Fig. 1G–I), indicating a transition block from DN3 to DN4 in Gcn5 conditional KO mice. In addition, although there is a statistically significant increase in the percentage of the CD25+Foxp3+ regulatory T cell (Treg) population in thymus (Fig. 1J, 1K), absolute Treg numbers are not affected by Gcn5 gene deletion (Fig. 1K). Therefore, Gcn5 functions appear to be important in regulating T cell development at the stage of DN3 to DN4 transition.
Analysis of T cell development in Gcn5 conditional KO mice. (A and B) CD44−CD62L+ naive T cells isolated from the spleen of WT and Gcn5 conditional KO mice were sorted. Gcn5 protein expression was determined by Western blotting (top panel) using β-Actin as a loading control (bottom panel) (A), and its mRNA was analyzed by real-time PCR (B). (C–K) Cells from the thymus of WT and Gcn5−/− mice were isolated (C). Single-cell suspension of thymocytes from the WT and Gcn5 conditional KO mice were analyzed by flow cytometry for their surface expression of CD4 and CD8 (C). The percentages (D) and absolute numbers (E) of CD4/CD8 DN, DP, and SP populations are shown. The DN populations were gated, and their expression of CD44 and CD25 were further analyzed (F). The percentages (G) and absolute numbers (H) of cells at DN1-DN4 stages are indicated. The CD4 SP cells were gated and the Foxp3+CD25+ Treg populations were analyzed (I). The percentages and total numbers of Foxp3+CD25+ Tregs are indicated (K). (L–N) Single-cell suspension of splenocytes from WT and Gcn5 KO mice were isolated. The percentages and numbers of CD4 and CD8 T cells (L) as well as Tregs (M) were analyzed. The absolute numbers of CD4, CD8, and Tregs in the spleen of seven pairs of WT and Gcn5-null mice are indicated (N). Student t test was used for the statistical analysis. *p < 0.01, **p < 0.005, ***p < 0.001.
As a consequence of the impaired T cell development, Gcn5 gene deletion caused a dramatic reduction in both CD4+ and CD8+ T cells compared with control WT littermates (Fig. 1L, 1N). In addition, there was an ∼2- to 3-fold increase in the percentage of CD25+Foxp3+ Tregs in spleens of Gcn5f/fLck-Cre+ mice (Fig. 1M, 1N), but the absolute Treg numbers were not altered (Fig. 1N), suggesting that the reduction in Treg percentage is likely a consequence of the CD4+ T cell decrease in the spleen of Gcn5 conditional KO mice.
Gcn5 gene deletion impairs T cell activation
Importantly, upon in vitro anti-CD3 plus anti-CD28 stimulation, purified naive CD4+ T cells from the Gcn5f/fLck-Cre+ mice showed a significant reduction in proliferation as determined by CFSE dilution (Fig. 2A) as a well as [3H]thymidine incorporation (Fig. 2B). Similar to the reduced T cell proliferation, the production of IL-2 by Gcn5-null CD4+ T cell is impaired (Fig. 2C). This reduction in T cell growth is unlikely due to elevated apoptosis because Annexin V–positive cell populations in WT and Gcn5-null T cells were comparable (Fig. 2D). IL-2 is a growth factor of activated T cells (34); it is possible that the impaired T cell growth is due to the reduced IL-2 production by Gcn5-null T cells. To test this, we analyzed T cell proliferation in the presence of exogenous recombinant murine IL-2. However, addition of IL-2 failed to fully rescue Gcn5-null T cell proliferation, implying that Gcn5 positively regulates T cell growth independent of IL-2.
The role of Gcn5 in T cell activation in vitro. Naive CD4 T cells were isolated from the spleens of WT and Gcn5 conditional KO mice. (A–D) T cells were cultivated with anti-CD3, anti-CD3 plus anti-CD28, or anti-CD3 plus anti-CD28 (1 μg/ml each) and IL-2 (2 ng/ml) for 3 d, and their proliferation was analyzed by either CFSE dilution (A) or [3H]thymidine incorporation (B). The IL-2 production in the culture supernatants was determined by ELISA (C). Cell viability was analyzed by propidium iodide and Annexin V staining (D). (E–G) Naive CD4 T cells were isolated from Gcn5loxp/loxpESR-CreTG and control WT littermates, and cultivated with IL-7 and 10 μg/ml tamoxifen for 7 d. Gcn5 expression was determined by Western blotting using Tubulin as a loading control (E). The tamoxifen cultivated cells were stained with CFSE and stimulated with anti-CD3 plus anti-CD28 for 3 d. T cell proliferation (F) and IL-2 production (G) were determined by flow cytometry. Error bars represent data from three independent experiments (mean ± SD). (H–J) Naive T cells from CD45.1 congenic WT mice and CD45.2 Gcn5 conditional knockout mice were mixed at a 1:1 ratio and adoptively transferred into T cell null (TCRβδ KO) mice (H). Seven days after the recipients were euthanized and total CD3+ T cells were gated for CD45.1 and CD45.2 analysis (I, left panel). The gated CD45.1 WT and CD45.2 Gcn5 KO CD4 and CD8 T cells were further analyzed (I, right panel). Representative images (I) and data from six recipient mice are shown (J) (mean ± SD). The Student t test was used for statistical analysis. *p < 0.01, **p < 0.005.
Because Cre recombinase expression under the Lck promoter drives Gcn5 gene deletion at the CD4 and CD8 DN2 during T cell development (26), together with the fact that Gcn5 gene deletion appears to impair T cell development, we reasoned whether the impaired T cell activation is a consequence of the developmental defect. To rule out this possibility, we determined that Gcn5 intrinsically promotes T cell activation. As previously reported (35), tamoxifen treatment of naive CD4+ T cells from Gcn5loxp/loxpESR-CreTG mice led to a complete deletion of Gcn5 as confirmed by Western blotting analysis (Fig. 2E). Notably, in vitro deletion of Gcn5 gene expression in vitro by tamoxifen treatment of mature CD4+ T cells from Gcn5loxp/loxpESR-CreTG mice resulted in a dramatic reduction in CD4+ T cell proliferation, as well as their IL-2 production (Fig. 2F, 2G). These results indicate that Gcn5 is a cell-autonomous positive regulator for T cell activation in vitro.
Furthermore, although the T cell developmental defect of Gcn5 conditional knockout mice is likely responsible for the reduced CD4 and CD8 T cells in their peripheral lymphoid organs, the possibility that Gcn5 is involved in regulating T cell homeostasis proliferation cannot be fully excluded. To test this hypothesis, we used an adoptive transfer approach and analyzed the homeostatic proliferation of T cells. Equal numbers of CD45.1 WT and CD45.2 Gcn5-null naive T cells (mixed at 1:1 ratio) were adoptively transferred into T cell–null mice (Fig. 2H). Seven days after adoptive transfer, CD4 and CD8 T cells were analyzed by flow cytometry. As indicated in Fig. 2I and 2J, both the percentage and the number of CD8+, but not CD4+, Gcn5 knockout T cells were reduced compared with the CD45.1 WT T cells in the same recipient, suggesting that, although Gcn5 gene deletion did not affect CD4 T cell homeostasis proliferation, Gcn5 is indispensable for CD8 T cell homeostasis and/or survival.
Gcn5 gene deletion does not affect the cell surface receptors of T cell activation
One possibility is that Gcn5 regulates T cell activation through altering the expression levels of cell surface TCR and costimulatory receptors. However, the cell surface expression levels of both TCRβ and the costimulatory receptor CD28 were comparable between WT and Gcn5-null CD4 T cells (Fig. 3A). Consistent with our observation of increased Treg percentages in the spleen, we detected a similar increase in the percentages of CD25+ CD4 T cells in the spleen. However, after a 24-h stimulation with anti-CD3 and anti-CD28, the expression levels of both CD25 and CD69 were comparable between WT and Gcn5-null CD4 T cells (Fig. 3B). Moreover, because T cell–specific Gcn5 gene deletion results in an ∼50% reduction in T cell population, we reasoned whether this lymphopenic phenotype leads to chronic T cell activation and consequently impairs T cell activation. However, the cell surface expression profiles of CD44 and CD62L on CD4 T cells were unaltered by Gcn5 deficiency (Fig. 3C). In addition, flow cytometry analysis did not detect any changes in the expression of IL-7Rα and IL-15Rα, both of which are involved in T cell homeostasis proliferation, on T cells (gated on the CD3ε+ population) in the spleen of Gcn5 conditional knockout mice (Fig. 3D, 3E). Therefore, these results indicated that Gcn5 regulates T cell activation and homeostasis proliferation is not through altering the cell surface receptor expression.
The expression of cell surface receptors on Gcn5-null T cells. Single-cell suspensions were prepared from spleens of WT and Gcn5 KO mice. (A) The expression levels of TCRβ and CD28, (B) CD25 and CD69, and (C) CD44 and CD62L on the gated CD4+ T cells were analyzed by flow cytometry. (D and E) Total naive T cells were isolated and cultivated with or without anti-CD3 plus anti-CD28 (1 μg/ml each) for 24 h, and the expression levels of IL-7Rα and IL-15Rα were analyzed. Representative images from six (A–C) or three (D and E) pairs of mice are shown.
Gcn5 gene deletion impairs T cell differentiation in vitro
We then analyzed the role of Gcn5 in CD4+ T cell differentiation by cultivating naive CD4+ T cells under Th1, Th2, Th17, and Treg polarization conditions as reported previously (31). As shown in Fig. 4A, the differentiation of IFN-γ–producing Th1 cells was largely reduced when Gcn5-null CD4+ T cells were polarized, indicating a positive role of Gcn5 in Th1 differentiation. In contrast, the differentiation of Gcn5-null CD4+ T cells into Th2 lineage was not inhibited. We also detected a slight but statistically significant reduction in both IL-17A+ and IL-17F+ Th17 cells by Gcn5 gene deletion, whereas Treg polarization was not decreased upon Gcn5 gene deletion (Fig. 4A). Analysis of cytokine production levels in culture supernatants by ELISA further confirmed that Gcn5 gene deletion leads to impaired IFN-γ and IL-17A, but not IL-4, production by CD4 T cells (Fig. 4B). Therefore, our studies indicate that Gcn5 positively regulates Th1 and Th17 without affecting Treg or Th2 differentiation.
Gcn5 is required for Th1 and Th17 differentiation in vitro. CD4 T cells were cultivated under Th1, TH2, Th17, or Treg polarization conditions for 5 d. (A) The production of cytokines and transcription factor Foxp3 were examined by intracellular staining. (B) The production of each cytokine in the culture supernatant was determined by ELISA. (C and D) The transcription factors for Th1 T-bet and Th17 ROR-γT (C), as well as the surface cytokine receptors (Th1: IL-12R and Th17: IL-6R, TGF-R, and IL-23R) (D) were determined by flow cytometry. (E–G) Naive CD4 T cells were stained with CFSE and cultivated under Th1 or Th17 polarization condition for 5 d. The production of each cytokine and T cell proliferation were analyzed by flow cytometry (E). The averages of cytokine production cells at each division are shown (F). The average MFIs are shown (G). Error bars represent data from three independent experiments (mean ± SD). The Student t test was used for statistical analysis. *p < 0.01, **p < 0.005, ***p < 0.001.
CD4 T cell differentiation is controlled by lineage-specific transcription factors. We determined whether Gcn5 promotes Th1 and Th17 differentiation by modulating the expression of their lineage-specific transcription factors T-bet and ROR-γT. Indeed, intracellular analysis detected a significant reduction in both T-bet and ROR-γT expression (Fig. 4C). These results indicate that Gcn5 regulates Th1 and Th17 differentiation through promoting T-bet and ROR-γT expression, respectively. After clonal expansion, and depending on the cytokine environment, CD4+ T cells differentiate into a variety of effector subsets, including Th1 cells and Th2 cells, the more recently defined Th17 cells, and induced Tregs (36). Therefore, it is possible that Gcn5 deficiency impairs CD4 T cell differentiation through altering the cell surface expression of cytokine receptors. However, Gcn5 deficiency appears not to regulate Th1 and Th17 differentiation through affecting their surface cytokine receptor expression, because the expression levels of cytokine receptors, IL-12 on Th1, IL-6R, TGF-β receptor, and IL-23R on Th17, were indistinguishable between WT and Gcn5-null T cells (Fig. 4D).
To further determine whether the defect in Th1 and Th17 cytokine production by Gcn5-null CD4 T cells is due to the impaired proliferation, we analyzed the cell cycle–based production of IFN-γ and IL-17 by Gcn5 KO CD4 T cells. The overall IFN-γ and IL-17 production by Gcn5-null CD4 T cells, as well as their cell cycle division, were reduced (Fig. 4E–G), confirming our initial observation that Gcn5 functions are required for the optimal T cell proliferation and cytokine production. Importantly, Gcn5 KO T cells produced markedly less IFN-γ and IL-17 compared with those of WT CD4 T cells even at the same cycle of cell division (Fig. 4E, 4F). Under Th17 polarization condition with TGF-β in the absence of IL-2, CD4 T cell proliferation is largely inhibited, which largely diminishes the reduction of Th17 proliferation by Gcn5 gene deletion (Fig. 4G). These results indicate that the reduced cytokine production by Gcn5 KO T cells is uncoupled with the cell cycle progression.
Gcn5 is required for Ag-specific T cell immune response in mice
To study the roles of Gcn5 in regulating Ag-specific T cell immune response in mice, we immunized Gcn5f/fLck-Cre+ mice and their littermate controls with chicken OVA emulsified with CFA at day 1, followed by a booster immunization with OVA/IFA at day 8. OVA-specific Ab production analyses indicate that T cell–specific Gcn5 gene deletion did not significantly affect OVA-specific IgM (Fig. 5A). In contrast, IgG1 levels were reduced at day 7, but not day 14, after immunization (Fig. 5B); this is likely due to that Gcn5 deletion did not affect Th2 differentiation. The production of IgG2a was impaired in mice with T cell–specific Gcn5 gene deletion (Fig. 5C). Consistent with the results that indicate Gcn5 deficiency dampens Th1, but not Th2, differentiation (Fig. 2), the percentages of IFN-γ–producing Th1 cells in the immunized mice were significantly reduced. In contrast, Th2 differentiation in the spleen of OVA-immunized Gcn5 conditional KO mice was not affected by Gcn5 gene deletion compared with that in WT littermate controls (Fig. 5D). These results indicate that Gcn5 positively regulates Th1-mediated Ag-specific immune response. In addition, the in vivo differentiation of IL-17–producing Th17 cells was slightly reduced by Gcn5 gene deletion (Fig. 5E), further supporting our in vitro observation that Gcn5 is possibly involved in Th17 differentiation.
The roles of Gcn5 in Ag-specific T cell immune response in vivo. Five pairs of 8- to 10-wk-old WT and Gcn5 conditional KO mice were immunized with OVA/CFA followed by a boosted immunization with OVA/IFA at day 8. (A–C) The levels of OVA-specific Abs including IgM (A), IgG1 (B), and IgG2a (C) in the serum were determined by ELISA. (D and E) Total splenocytes were stimulated with PMA (20 ng/ml) with ionomycin (500 ng/ml) and brefeldin A (10 μg/ml) for 4 h and then analyzed for IFN-γ–producing Th1, IL-4–producing Th2 (D), and IL-17–producing Th17 cells (E) by intracellular staining. (F–J) Total splenocytes were stained with CFSE and cultivated with OVA at each indicated concentration for 3 d. The proliferation of CD4 T cells was analyzed by flow cytometry (F). The production of IL-2 (G), IFN-γ (H), IL-4 (I), and IL-17 (J) was analyzed by ELISA. Error bars represent data from five pairs of mice (mean ± SD). Student t test was used for the statistical analysis. *p < 0.01, **p < 0.005.
When the total splenocytes from the immunized mice were cultivated with OVA Ag, significant reductions in both CD4+ T cell proliferation (Fig. 5F) and IL-2 production (Fig. 5G) were detected, indicating that Gcn5 functions are required for Ag-specific T cell activation. In addition, the production of both IFN-γ and IL-17, but not IL-4, in the culture supernatants was inhibited by Gcn5 gene deletion (Fig. 5H–J). Collectively, we conclude that Gcn5 is a positive regulator in T cell activation and Th1/Th17 differentiation in mice when challenged with specific Ags.
Gcn5 is recruited to il-2 promoter through NFAT
As an HAT, Gcn5 has been shown to play important roles in promoting gene transcription by modifying either specific transcription factors or histones at specific gene promoters. Because Gcn5 positively regulates T cell production of IL-2, we first asked whether Gcn5 is recruited onto the promoter region of il-2 in T cells. As indicated in Fig. 6A, ChIP analysis detected a background level of Gcn5 binding to the il-2 promoter in naive T cells. Stimulation with anti-CD3 or with anti-CD3 plus anti-CD28 dramatically enhanced Gcn5 association with the il-2 promoter in mouse primary CD4+ T cells. TCR stimulation appeared to be sufficient to enhance the recruitment of Gcn5 onto the il-2 promoter because addition of anti-CD28 did not further increase Gcn5 recruitment. These results indicate that stable association of Gcn5 to the il-2 promoter requires TCR signaling. Gcn5 is recruited to target promoters through interactions with transcription factors. NFAT, AP-1, and NF-κB are critical for il-2 gene transcription (37), so we tested whether Gcn5 is recruited to the il-2 promoter through interactions with any of these transcription factors by treatment of CD4+ T cells with specific inhibitors. Notably, the calcineurin-specific inhibitor CsA largely blocked Gcn5 recruitment to the il-2 promoter (Fig. 6B). The JNK inhibitor SP60025 that specifically suppresses AP-1 transcriptional activity also suppressed Gcn5 binding to the il-2 promoter, albeit to a lesser degree than CsA. In contrast, the NF-κB inhibitor JHS-23 did not affect Gcn5 association with the il-2 promoter in T cells (Fig. 6B). Therefore, the recruitment of Gcn5 onto the il-2 promoter likely occurs mostly through NFAT activation.
Gcn5 interacts with NFAT in T cells. (A and B) Naive CD4 T cells were stimulated with anti-CD3 (1 μg/ml) or with anti-CD3 plus anti-CD28 (1 μg/ml) (A), or further with DMSO or each of the inhibitors as indicated (B) for 6 h. The binding of Gcn5 to the il-2 promoter was analyzed by ChIP. (C) Flag-Gcn5 was cotransfected with HA-NFAT1 into HEK293 cells. Total cell lysates were subjected to co-IP with anti-HA, and the interaction of NFAT with Gcn5 was determined by Western blotting with anti-Flag Ab (top). The same membrane was reprobed with anti-HA (middle). Gcn5 expression in the whole cell lysates was confirmed by Western blotting (bottom). (D and E) Naive T cells were stimulated with anti-CD3 or anti-CD3 plus anti-CD28 (D) or further with 100 ng/ml CsA (E) for 6 h. The interaction of NFAT1 and Gcn5 was determined by co-IP and Western blotting. (F) Naive T cells were stimulated with anti-CD3 plus anti-CD28 for 16 h in the presence or absence of CsA. Subcellular fractionation was performed, and the protein levels in cytoplasmic (C) and nuclear (N) fractions were determined by Western blotting. *p < 0.01, **p < 0.005.
We then determined whether Gcn5 physically interacts with NFAT. NFAT family proteins have five members, and both NFAT1 and NFAT2 play predominant roles in T cell activation (38). We determined whether NFAT1 (also known as NFATc2) interacts with Gcn5 in transiently transfected HEK293 cells. As shown in Fig. 6C, Western blotting detected Gcn5 in the anti-HA (NFAT1) immunoprecipitates from lysates of HA-NFAT1 transfected cells, but not in lysates from nontransfected controls. We then determined whether endogenous Gcn5 interacts with NFAT1 in mouse primary T cells. When the lysates of naive CD4+ T cells were subjected to coimmunoprecipitation (co-IP) with anti-NFAT1 Ab, a weak Gcn5 band was detected in the anti-NFAT1 immunoprecipitates from the lysates of naive CD4 T cells (Fig. 6D). Anti-CD3 stimulation for 6 h significantly enhanced NFAT1-Gcn5 interaction, but addition of anti-CD28 Ab did not further enhance this interaction, indicating that TCR signaling alone is sufficient to promote NFAT1 interaction with Gcn5 (Fig. 6D). Because CsA, which blocks NFAT1 nuclear translocation, inhibited the Gcn5 recruitment to the il-2 promoter (Fig. 6B), we tested whether CsA inhibits NFAT1 interaction with Gcn5. In fact, CsA treatment dramatically attenuated NFAT1 interaction with Gcn5 in mouse primary T cells (Fig. 6E), suggesting that Gcn5 binds to the il-2 promoter through NFAT1 interaction.
We noticed that CsA treatment did not result in a full block of Gcn5 binding to il-2 promoter in T cells (Fig. 6B); this is likely because CsA treatment only partially blocked NFAT nuclear translocation. Indeed, as determined by subcellular fractionation analysis, although with a significant reduction, a substantial amount of NFAT protein was detected in the nuclei of CsA-treated T cells (Fig. 6F). Interestingly, CsA treatment did not alter Gcn5 nuclear localization (Fig. 6F). However, treatment of T cells with CsA largely inhibited the Gcn5 recruitment onto the il-2 promoter as documented by ChIP assay (Fig. 6B). These results suggest that the recruitment of Gcn5 onto the il-2 promoter, but not its overall nuclear localization, is mediated by NFAT.
Gcn5 catalyzes H3K9 acetylation at the il-2 promoter during T cell activation
Although NFAT acetylation has not been reported, Gcn5-mediated acetylation of transcription factors such as NF-κB often enhances their transcriptional activity (12, 14, 19, 39), so we analyzed whether Gcn5 catalyzes NFAT1 acetylation to promote il-2 gene transcription in T cells. However, NFAT1 acetylation was not detected either in HEK293 cells when Gcn5 was coexpressed or in T cells after CD3/CD28 stimulation (Fig. 7A, 7B), suggesting that Gcn5-regulated il-2 gene transcription does not likely involve NFAT1 acetylation.
Gcn5 regulates IL-2 production in T cells through NFAT. (A and B) Naive T cells were isolated from WT and Gcn5 conditional KO mice and activated by anti-CD3 plus anti-CD28 for 16 h. The level of acetylated H3K9 was determined by Western blotting (top). The expression levels of total H3 (middle) and Gcn5 (bottom) were confirmed by Western blotting. Representative image (A) and the average levels of Ac-H3K9 from four independent experiments (B) are shown. (C–F) Naive T cells from WT and Gcn5 KO mice were activated with anti-CD3 plus anti-CD28 (C–E) or further treated with CsA (F). The levels of total acetyl-H3K9 were determined by Western blotting with acetyl-H3K9–specific Abs (top panel). The protein levels of H3 (middle) and Gcn5 (bottom) were analyzed by Western blotting (C). The acetyl-H3K9 levels were normalized by its total H3 protein levels, and averages from four independent experiments are shown (D–F). The acetyl-H3K9 levels at il-2 promoter were determined by ChIP assay. Error bars represent data from five pairs of mice (mean ± SD) (D–F). Student t test was used for the statistical analysis. *p < 0.01, **p < 0.005. (G) Flag-Gcn5 was cotransfected with HA-NFAT2 into HEK293 cells. Total cell lysates were subjected to co-IP with anti-HA, and the interaction of NFAT2 with Gcn5 was determined by Western blotting with anti-Flag Ab (top). The same membrane was reprobed with anti-HA (middle). Gcn5 expression in the whole cell lysates was confirmed by Western blotting (bottom).
Gcn5 positively regulates gene transcription by catalyzing histone H3K9 acetylation (40). To determine whether Gcn5 promotes T cell activation through H3K9 acetylation, we compared H3K9 acetylation between WT and Gcn5-null T cells. As shown in Fig. 7C, only a background level of H3K9 acetylation could be detected in naive T cells. Stimulation with anti-CD3 and anti-CD28 induced a significant increase in H3K9 acetylation in WT T cells. In contrast, a slight but reproducible decrease in the levels of H3K9 acetylation was detected in Gcn5-null T cells (Fig. 7C, 7D), suggesting that Gcn5 is involved in regulating global H3K9 acetylation in T cells. We then speculated that Gcn5 might enhance IL-2 production through catalyzing H3K9 acetylation at the il-2 promoter. Indeed, ChIP analysis detected a significant upregulation of H3K9 acetylation at the il-2 promoter in WT T cells after TCR/CD28 stimulation. Importantly, loss of Gcn5 functions largely abolished H3K9 acetylation at the il-2 promoter, indicating that Gcn5 is responsible for H3K9 acetylation at the il-2 promoter during T cell activation (Fig. 7E). To further determine whether Gcn5 catalyzes H3K9 acetylation through interactions with NFAT, we analyzed the effect of calcineurin inhibitor CsA on H3K9 acetylation at the il-2 promoter. As expected, CsA pretreatment largely abolished H3K9 acetylation at the il-2 promoter in WT, but not Gcn5-null T cells (Fig. 7F). In addition to NFAT1, we also detected an interaction of Gcn5 with NFAT2 in transiently transfected cells (Fig. 7G), implying that other NFAT family proteins, at the least NFAT2, can also be recruited onto the il-2 promoter through Gcn5 interaction. Collectively, our study indicates that Gcn5 is recruited to the il-2 promoter through NFAT to catalyze H3K9 acetylation during T cell activation.
Genetic suppression of Gcn5 partially protects mice from autoimmune disease
The fact that Gcn5 positively regulates T cell activation implies Gcn5 as a potential therapeutic target in autoimmune disease therapy. We then used an MOG-induced EAE model and tested whether T cell–specific Gcn5 depletion protects mice from autoimmune response. As shown in Fig. 8A, after being immunized with MOG35–55 peptides, WT mice gradually developed demyelination disease beginning at day 9 after initial immunization. Clinical symptoms peaked around day 17 with an average clinical score of 3.5. In contrast, disease onset was delayed by at least 3 d in Gcn5 conditional KO mice, and their clinical symptoms were more modest (Fig. 8A), clearly indicating that genetic depletion of Gcn5 in T cells partially protects mice from autoimmune disease. Further analysis confirmed that the MOG-specific T cell immune response in Gcn5 conditional KO mice was significantly reduced, because MOG-specific proliferation of CD4+ T cells in the spleen was largely inhibited by Gcn5 gene deletion (Fig. 8B). Intracellular cytokine staining analysis showed that Gcn5 deficiency dramatically inhibited production of IL-2, IFN-γ, and IL-17 by CD4+ T cells in the spleen during EAE (Fig. 8C). Consistent with our results from in vitro analyses, the percentages of IL-4–producing Th2 cells were not inhibited in the spleen of Gcn5 KO mice (Fig. 8C). Collectively, our study shows that Gcn5 is a positive regulator in T cell activation, and that suppression of Gcn5 functions lessens the effects of autoimmune disease in this mouse model. To further support this conclusion, we demonstrated that the infiltrated T cells, both CD4 and CD8, were significantly reduced in MOG-immunized Gcn5 KO mice (Fig. 8D, 8E). Intracellular analysis confirmed that the production of pathogenic cytokines including IFN-γ and IL-17 were both significantly reduced in the infiltrated CD4 T cells isolated from the CNS of Gcn5 KO mice compared with the WT controls. Consistent with our initial observation, the percentage of CNS-infiltrated Foxp3+ Tregs in Gcn5 KO mice was higher (Fig. 8F). Therefore, these studies imply that Gcn5 is a potential therapeutic target for multiple sclerosis treatment.
Analysis of MOG-induced EAE in Gcn5 conditional KO mice. EAE was induced in six pairs of WT and Gcn5 KO mice at the age of 9 wk. (A) The clinical symptoms were scored every 3 d. The average clinical scores are indicated. (B) Total splenocytes were stained with CFSE and cocultivated with MOG35–55 peptide for 3 d. CD4+ T cells were gated and the cell proliferation was determined by flow cytometry for CFSE dilution. (C) Splenocytes from MOG-immunized mice at day 17 after immunization during the peak of the disease were stimulated with PMA and ionomycin for 4 h. The productions of IL-2, IFN-γ, IL-4, and IL-17 by CD4 T cells were determined by intracellular staining followed by flow cytometry analysis. Representative images from six mice are shown. (D–F) MOG-immunized mice were euthanized at day 17 after immunization. Infiltrated lymphocytes in the CNS were isolated. T cells were determined by flow cytometry for their expression of CD3 (D, top), and the gated CD3+ T cells were further characterized by their expression of CD4 and CD8 (D, bottom). The average T cell numbers from six pairs of mice are shown (E). The levels of Th1, Th17, and Tregs were analyzed by intracellular staining (F). Student t test was used for the statistical analysis. *p < 0.01, **p < 0.005.
Discussion
Based on our discoveries, we propose a model for Gcn5 regulation of T cell activation: when stimulated by specific Ags, TCR signaling induces the nuclear translocation of NFAT. NFAT then binds to and recruits Gcn5 to the il-2 promoter, where it promotes H3K9 acetylation to activate il-2 gene transcription. Suppression of Gcn5 functions inhibits T cell activation, suggesting that Gcn5 may provide a potential therapeutic target for treatment of autoimmune disease. Therefore, GCN5 is a positive regulator of T cell activation and CD4 T cell differentiation into Th1 and Th17. This conclusion is supported by the following observations: 1) the targeted Gcn5 deletion specifically in T cells partially blocks T cell development at the stage from DN2 to DN3 transition; 2) loss of Gcn5 function attenuates T cell activation in vitro and in vivo; 3) Gcn5 is recruited onto the il-2 promoter through NFAT interaction to modify H3K9 acetylation; and 4) T cell–specific Gcn5 gene deletion largely protects mice from MOG-induced EAE.
Inhibitors of HAT have been studied for their potential in autoimmune therapy (41). For example, several small molecules that suppress the acetyltransferase activity of p300 show promising effects in suppressing T cell activation and autoimmune diseases in small animal models (42). However, because p300 is critical for Treg function, suppression of p300 can impair Treg functions (43, 44). Our discoveries in this study suggest that Gcn5 inhibition is an attractive therapeutic target for autoimmune disease treatment or immune suppression to prolong the graft survival in organ transplantation, because Gcn5 suppression impairs T cell clonal expansion, inhibits IL-2 production and Th1, Th17 differentiation, but has not effect on Treg and Th2 differentiation. Small-molecule inhibitors that specifically suppress Gcn5 acetyltransferase catalytic activity have been recently reported (45, 46); our laboratory is currently testing the effects of Gcn5 inhibitors in T cell activation, differentiation, and autoimmune therapy.
Transcription factor–dependent recruitment of epigenetic modification enzymes to specific promoters is a common molecular mechanism in regulating gene transcription. We have recently shown that the histone deacetylase Sirt1 is recruited to the promoter of Bclaf1 to suppress Bclaf1 gene transcription by attenuating H3K27 acetylation through binding the NF-κB transcription factor RelA (47). Similarly, it has been shown that RelA recruits Sirt6 to suppress NF-κB target genes (48). In this article, we identify Gcn5 as an interactor of NFAT1 in T cells, and their interaction requires TCR signaling. However, expression of Gcn5 failed to promote NFAT1 acetylation, and NFAT1 acetylation was not detectable in T cells even after TCR/CD28 stimulation. This excludes the possibility that Gcn5 regulates T cell activation through catalyzing NFAT1 acetylation. Notably, Gcn5 nuclear translocation is independent of NFAT because the calcineurin-specific inhibitor CsA did not alter nuclear Gcn5 protein levels but largely blocked Gcn5 recruitment onto the il-2 promoter. Importantly, Gcn5 is recruited onto the il-2 promoter through binding to NFAT1 to promote H3K9 acetylation for IL-2 production. Stable histone acetylation on il-2 promoter has been suggested as a critical epigenetic regulation of T cell activation (49). Therefore, Gcn5 is a critical epigenetic factor in promoting il-2 gene transcription through NFAT during T cell activation.
Gcn5 appears to be involved in regulating T cell function at multiple stages because genetic Gcn5 deletion partially blocks T cell development at the stage from DN3 to DN4 transition and impairs T cell clonal expansion, both of which rely on vigorous cell proliferation. Because of limiting cell numbers, it is technically challenging to investigate whether Gcn5 regulates T cell development through its interaction with NFAT during the transition from the DN2 to the DN3 stage (41, 42). However, we speculate that interactions between Gcn5 and NFAT will be at least partially involved in regulating DN3 transition to DN4 because genetic deletion of NFAT1 also impairs T cell development at this stage (43). Our studies show that GCN5 regulates T cell proliferation at both unpolarization condition and during Th1 and Th17 differentiation. However, we also noticed that the reduction in proliferation during Th17 polarization condition is subtle compared with Th1; this is possibly due to the modest T cell expansion during Th17 polarization with TGF-β in the absence of IL-2.
In addition to development and activation, Gcn5 is involved in CD4+ T cell differentiation, because both Th1 and Th17 polarizations are impaired by Gcn5 gene deletion. After clonal expansion, and depending on the cytokine environment, CD4+ T cells differentiate into a variety of effector subsets, including Th1 cells and Th2 cells, the more recently defined Th17 cells, and induced Tregs (36). However, the expression levels of cytokine receptors critical for Th1 differentiation, that is, IL-12R or Th17 differentiation including the TGF-β receptor, IL-6R, and IL-23R, were not altered by Gcn5 gene deletion. Importantly, a dramatic reduction in the expression of the lineage-specific transcription factor for Th1, T-bet, and Th17, ROR-γT, was detected in Gcn5-null CD4 T cells. Further studies are needed to elucidate the molecular mechanisms underlying how Gcn5 promotes T-bet and ROR-γT expression. Despite loss of Gcn5 in CD4 T cells did not affect in vitro Treg polarization, there is a 2- to 3-fold increase in Treg percentage in Gcn5 conditional KO mice. This is likely due to the relative reduction of naive CD4 T cells because the absolute numbers of CD25+Foxp3+ Tregs are comparable between WT and Gcn5 conditional KO mice. Moreover, Gcn5 appears to be dispensable for Th2 differentiation, making this molecule an ideal therapeutic target for autoimmune disease. To support this notion, we show in this article that T cell–specific Gcn5 gene deletion partially protected mice from MOG-induced EAE, an experimental model for human multiple sclerosis.
Disclosures
The authors have no financial conflicts of interest.
Footnotes
This work was supported by National Institutes of Health Grants R01 AI079056 (to D.F.) and R01 GM067718 (to S.Y.R.D.).
The online version of this article contains supplemental material.
Abbreviations used in this article:
- ChIP
- chromatin immunoprecipitation
- co-IP
- coimmunoprecipitation
- CsA
- cyclosporine A
- DN
- double-negative
- DN2
- DN stage 2
- DN3
- DN stage 3
- DN4
- DN stage 4
- DP
- double-positive
- EAE
- experimental autoimmune encephalomyelitis
- Gcn5
- control nonrepressed–protein 5
- HAT
- histone acetyltransferase
- H3K9
- lysine residue 9 of histone H3
- KO
- knockout
- MOG
- myelin oligodendrocyte glycoprotein
- SP
- single-positive
- Treg
- regulatory T cell
- WT
- wild-type.
- Received February 22, 2016.
- Accepted March 20, 2017.
- Copyright © 2017 by The American Association of Immunologists, Inc.