MATERIALS AND METHODS

Mice.
CbfbF/F CD4-cre and Foxp3GFP mice have previously been described (Bettelli et al., 2006; Naoe et al., 2007). Cd45.1 and Rag2−/− mice were purchased from Jackson ImmunoResearch Laboratories and Taconic, respectively. For the in vivo Foxp3 conversion assay, naive CD4+ T cells from Cbfb CD4-cre or control mice (harboring a Foxp3-IRES-GFP allele) were adoptively transferred into Rag-deficient mice. 5 × 106 cells were used per transfer. 6 wk later, TCRβ+CD4+ gated cells from the spleen, mesenteric lymph node (MLN), and lamina propria of the small intestine were analyzed for Foxp3-GFP expression. All analyses and experiments were performed on animals at 6–8 wk of age. Animals were housed under specific pathogen–free conditions at the animal facility of the Skirball Institute, and experiments were performed in accordance with approved protocols for the New York University Institutional Animal Care and Usage Committee.

Isolation of PBMCs, CD4+ T cells, and culture conditions.
Human PBMCs were isolated by Ficoll (Biochrom) density gradient centrifugation and CD4+ T cells were then isolated using the Dynal CD4+ Isolation kit (Invitrogen) according to the manufacturer's instructions. The purity of CD4+ T cells was initially tested by flow cytometry and was ≥95%. Cells were stimulated with the following combination of mAbs to T cell surface molecules (Meiler et al., 2008): anti-CD2 (clone 4B2 and 6G4; 0.5 µg/ml), anti-CD3 (clone OKT3; 0.5 µg/ml), and anti-CD28 mAb (clone B7G5; 0.5 µg/ml; all from Sanquin) and cultured in serum-free AIM-V medium (Life Technologies) with the addition of 1 nmol/liter IL-2 (Roche). TGF-β (R&D Systems) was used at 5 ng/ml, if not stated otherwise. A combination of PMA (25 ng/ml) and ionomycin (1 mg/ml; Sigma-Aldrich) was used.
Human CD4+ CD127− CD25high and CD4+ CD127+ CD25neg cells were purified by flow cytometry using anti-CD127, anti–CD25-PC5 (Beckman Coulter), and anti–CD4-FITC antibodies (Dako).
Mouse naive (CD62Lhi44lo25−) CD4+ T cells were purified by flow cytometry and activated in vitro with 5 µg/ml plate-bound anti-CD3 and 1 µg/ml soluble anti-CD28 antibodies (eBioscience) in RPMI supplemented with 10% FCS, 5 mM β-mercaptoethanol, and antibiotics. Neutralizing anti–IFN-γ and anti–IL-4 mAbs (BD) were used at 1 µg/ml concentrations when indicated.

In vitro T cell differentiation.
CD4+ CD45RA+ magnetically sorted (CD45RO depletion with AutoMACS; Miltenyi Biotec) cells were stimulated with immobilized plate-bound anti-CD3 (1 µg/ml; OKT3; IgG1) and anti-CD28 (2 µg/ml). For Th1 differentiation conditions, cells were stimulated with the following: 40 ng/ml IL-2, 5 µg/ml anti–IL-4, and 25 ng/ml IL-12 (R&D Systems). For Th2 conditions, cells were stimulated with the following: 40 ng/ml IL-2, 25 ng/ml IL-4, and 5 µg/ml anti–IL-12 (R&D Systems). For T reg cell conditions, cells were stimulated with the following: 40 ng/ml IL-2, 5 ng/ml TGF-β, 5 µg/ml anti–IL-12, 5 µg/ml anti–IL-4. For Th17 conditions, cells were stimulated with the following: 40 ng/ml IL-2, 20 ng/ml IL-6, 5 ng/ml TGF-β, 20 ng/ml IL-23 (Alexis Biochemicals Corp.), 10 ng/ml IL-1β, 5 µg/ml anti-IL-4, and 5 µg/ml anti–IL-12 were used. Proliferating cells were expanded in medium containing IL-2. The cytokine profile of these cells demonstrated that IFN-γ is the predominant cytokine in Th1 cells, IL-4 and IL-13 in Th2 cells, and IL-17 in Th17 cells (Akdis et al., 2000; Burgler et al., 2009).

Immunohistochemistry.
Human tonsils were obtained from tonsillectomy samples of hypertrophic and obstructive tonsils without a current infection. Ethical permission was obtained from Cantonal Ethics Commission, and informed consent was obtained from patients. Paraformaldehyde-fixed tonsil cryosections were stained with unconjugated rabbit IgG polyclonal antibody to human RUNX1 (Santa Cruz Biotechnology, Inc.) or unconjugated mouse IgG1 mAb to human RUNX3 (Abcam). After a washing step, the sections were stained with the corresponding secondary antibodies. RUNX1-binding antibodies were detected by using Alexa Fluor 633–conjugated goat anti–rabbit IgG and RUNX3-binding antibodies were detected by using Alexa Fluor 532–conjugated goat anti–mouse IgG1. Afterward, the sections were washed and in the case of RUNX3 staining a blocking step with an unconjugated mouse IgG1 mAb was used. Finally, the sections were stained with Alexa Fluor 488–conjugated mouse IgG1 mAb to human FOXP3 (eBioscience) or the corresponding isotype control. Tissue sections were stained with DAPI for the demonstration of nuclei and mounted with Prolong antifade (Invitrogen). Images were acquired and analyzed using the confocal microscope DMI 4000B and the TCS SPE system (both from Leica).

In vitro suppression assays.
Mouse Foxp3+ CD4+ CD8− T cells were FACS purified based on GFP expressed from a Foxp3-IRES-GFP knock-in allele (Foxp3GFP). Naive (CD62Lhi44lo25−) CD4+ T cells (effectors) were FACS-purified from Cd45.1 mice, then loaded with 5 µM CFSE (Invitrogen). Total splenocytes from C57BL/6 mice inactivated with 50 µg/ml mitomycin C (Sigma-Aldrich) for 45 min were used as APCs. A total of 4 × 105 CD4+ cells (CFSE-loaded CD25− plus Foxp3-GFP+) was mixed with 105 APCs + 1 µg/ml anti-CD3 mAb per well of a 96 well round bottom plate. Proliferation of the effector cells was analyzed by CFSE dilution. Apoptosis of the cells was investigated by annexin V staining and flow cytometry. Positive and negative control gates were made according to T cells cultured only in the presence of IL-2 without anti-CD3/28 stimulation.
Human naive CD4+ T cells were isolated by negative selection by MACS from PBMCs and either transfected with a scrambled siRNA or with a combination of RUNX1 and RUNX3 siRNA. Cells were then cultured under iT reg cell differentiating conditions and mixed with 2 × 105 autologous irradiated PBMCs that were used as APCs and autologous CFSE-labeled CD4+ T cells. T reg cell to responder cell ratio was 1:20, 1:10, and 1:5. To check the proliferation of the CD4+ T cells without suppression, no T reg cells were added in a control group. Cells were stimulated with 2.5 µg/ml anti-CD3 mAb, cultured in a 96-well plate and the proliferation of the effector cells was determined by analyzing the CFSE dilution by flow cytometry after 5 d of culture. Gating on the CD4+CFSE+ T cells enabled the exclusion of APCs and T reg cells.

Cloning of the FOXP3 promoter, construction of mutant FOXP3 promoter, and RUNX expression plasmids.
The FOXP3 promoter was cloned into the pGL3 basic vector (Promega Biotech) to generate pGL3 FOXP3 -511/+176 (Mantel et al., 2006). Site-directed mutagenesis for the three putative RUNX binding sites in the FOXP3 promoter region was introduced using the QuickChange kit (Stratagene), according to the manufacturer's instructions and confirmed by sequencing the DNA. The following primers and their complementary strands were used: foxp runx-333, forward 5′-CACTTTTGTTTTAAAAACTGTCCTTTCTCATGAGCCCTATTATC-3′; foxp runx-333 reverse 5′-GATAATAGGGCTCATGAGAAAGGACAGTTTTTAAAACAAAAGTG-3′; foxp runx-287 forward 5′-CCTCTCACCTCTGTCCTGAGGGGAAGAAATC-3′; foxp runx-287 reverse 5′-GATTTCTTCCCCTCAGGACAGAGGTGAGAGG-3′; foxp runx-53 forward 5′-GCTTCCACACCGTACAGCGTCCTTTTTCTTCTCGGTATAAAAG-3′; foxp runx-53 reverse 5′-CTTTTATACCGAGAAGAAAAAGGACGCTGTACGGTGTGGAAGC-3′.
The human RUNX1 fragment from the Addgene plasmid 12504 (Biggs et al., 2006) pFlagCMV2-AML1B was sub-cloned in the pEGFPN1 vector (Clontech Laboratories). The RUNX3 vector pCMV human RUNX3, which was a gift from K. Ito (Institute of Molecular and Cell Biology, Proteos, Singapore) was subcloned into pEGFPN1 vector (Clontech Laboratories; Yamamura et al., 2006).

Transfections and reporter gene assays.
T cells were rested in serum-free AIM-V medium overnight. 3.5 µg of the FOXP3 promoter luciferase reporter vector or a combination together with the RUNX1, RUNX3 pEGFPN1 vector, and 0.5 µg phRL-TK were added to 3 × 106 CD4+ T cells resuspended in 100 µl of Nucleofector solution (Lonza) and electroporated using the program U-15. After a 24-h culture in serum-free conditions and stimuli as indicated in the figures, luciferase activity was measured by the dual luciferase assay system (Promega) according to the manufacturer's instructions. PMA/ionomycin was used to stimulate the cells, because the transfection was only transient and the luciferase assay required a strong and fast stimulation of the cells. To evaluate the effect of overexpression of RUNX1 or RUNX3 on FOXP3 protein levels, CD4+ T cells were preactivated with 2 µg/ml phytohemagglutinin (Sigma-Aldrich) in serum-free AIM-V medium in the presence of 1 nmol/liter IL-2 (Roche) for 12 h, and then transfected with the vector pEGFPN1 containing the RUNX1 or RUNX3 fragment using the Nucleofector system (Amaxa Biosystems) and the program T-23. FOXP3 expression was evaluated by flow cytometry after 48 h of culture in AIM-V medium containing 1 nmol/liter IL-2.

RNA interference.
CD4+ or naive CD4+ T cells were resuspended in 100 µl of Nucleofector solution (Lonza) and electroporated with 2 µM siRNA using the Nucleofector technology program U-14 (Lonza). Five different Silencer or Silencer Select Pre-designed siRNAs for RUNX1 (Applied Biosystems) and three Silencer Pre-designed siRNAs for RUNX3 (Applied Biosystems) were tested, and the best was selected for all further experiments. The Silencer Negative Control #1 siRNA (Applied Biosystems) was used for normalization. Cells were then left unstimulated or were stimulated after 12 h with anti-CD2, anti-CD3, and anti-CD28. Cells were cultured in serum-free AIM-V medium with the addition of 1 nmol/l IL-2 (Roche). Cells were harvested for mRNA detection of the target genes after 24 h and for protein detection after 48 h.

RNA isolation and cDNA synthesis.
RNA was isolated using the RNeasy Mini kit (QIAGEN) according to the manufacturer's protocol. Reverse transcription of human samples was performed with reverse-transcription reagents (Fermentas) with random hexamers according to the manufacturer's protocol.

Real-time PCR.
PCR primers and probes were designed based on the sequences reported in GenBank with the Primer Express software version 1.2 (Applied Biosystems) as follows: FOXP3 forward primer, 5′-GAAACAGCACATTCCCAGAGTTC-3′; FOXP3 reverse primer, 5′-ATGGCCCAGCGGATGAG-3′; EF-1a forward primer, 5′-CTGAACCATCCAGGCCAAAT-3′; and EF-1a reverse primer, 5′-GCCGTGTGGCAATCCAAT-3′, as previously described (Mantel et al., 2007). GATA3 forward primer, 5′-GCGGGCTCTATCACAAAATGA-3′; and GATA3 reverse primer 5′-GCTCTCCTGGCTGCAGACAGC-3′ (Mantel et al., 2007). T-bet forward primer, 5′-GATGCGCCAGGAAGTTTCAT-3′; T-bet reverse primer, 5′-GCACAATCATCTGGGTCACATT-3′; RORC2 forward primer, 5′-CAGTCATGAGAACACAAATTGAAGTG-3′; and RORC2 reverse primer 5′-CAGGTGATAACCCCGTAGTGGAT-3′. The prepared cDNAs were amplified using SYBR green PCR master mix (Fermentas) according to the recommendations of the manufacturer in an ABI PRISM 7000 Sequence Detection System (Applied Biosystems).
RUNX1 and RUNX3 mRNA was detected by using TaqMan Gene Expression Assays from Applied Biosystems and used according to the manufacturer's instruction using TaqMan master mix using a 7000 real-time PCR system (Applied Biosystems). PCR amplification of the housekeeping gene encoding elongation factor (EF)-1α or by using the 18S rRNA Gene Expression Assay (Applied Biosystems) was performed to allow normalization between samples. Relative quantification and calculation of the range of confidence was performed using the comparative ΔΔCT method (Applied Biosystems). The percentage of FOXP3 mRNA in siRNA-mediated RUNX knockdown cells was calculated in relation to cells, which were transfected with scrambled control siRNA. Arbitrary units show the 2−(Δct) values multiplied by 10,000 incorporating the ct values of the gene of interest and the housekeeping gene.

Flow cytometry.
For analysis of human FOXP3 expression on the single-cell level, cells were first stained with the monoclonal CD4 mAb (Beckman Coulter), and after fixation and permeabilization, they were incubated with anti–human Foxp3-Alexa Fluor 488 antibody (BioLegend) based on the manufacturer's recommendations and subjected to FACS (EPICS XL-MCL; Beckman Coulter). A mouse IgG1 antibody (BioLegend) was used as an isotype control. Data were analyzed with the CXP software (Beckman Coulter). Cells were cultured with IL-2, and then left unstimulated or stimulated with anti-CD2/-CD3/-CD28 mAb.
Flow cytometry analyses of the mouse cells were performed on an LSRII (BD) and cell sorting was performed on a FACSAria (BD). All antibodies for these experiments were purchased from eBioscience or BD.

Western blotting.
For human RUNX1 and RUNX3 analysis on the protein level, 106 cells were lysed and loaded next to a protein-mass ladder (Invitrogen) on a NuPAGE 4–12% Bis-Tris gel (Invitrogen). The proteins were electroblotted onto a PVDF membrane (GE Healthcare). Unspecific binding was blocked with 3% milk in TBS Tween, and the membranes were subsequently incubated with a 1:1,000 dilution of rabbit anti-RUNX1 (ab11903; Abcam) or 1:200 dilution of rabbit anti-RUNX3 (H-50; Santa Cruz Biotechnology, Inc.) in blocking buffer containing 3% milk in TBS Tween overnight at 4°C. The blots were developed using an anti-rabbit IgG HRP-labeled mAb (Cell Signaling Technology) and visualized with a LAS-1000 gel documentation system (Fujifilm). To confirm sample loading and transfer membranes were incubated in stripping buffer and reblocked for 1 h and reprobed using anti-GAPDH (6C5; Ambion) and developed using an anti–mouse IgG HRP-labeled mAb (Cell Signaling Technology).

Pull-down assay.
HEK293T cells were transfected with RUNX1 or RUNX3 using the Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer's instruction. Cells were lysed by sonication in HKMG buffer (10 mM Hepes, pH 7.9, 100 mM KCl, 5 mM MgCl2, 10% glycerol, 1 mM DTT, 0.5% Nonidet P-40) containing a protease inhibitor cocktail (Roche Diagnostics). The cell lysate was precleared using streptavidin-agarose beads (GE Healthcare), incubated with biotinylated double-stranded oligonucleotides containing the wild-type or mutated RUNX binding sites, and polydeoxyinosinicdeoxycytidylic acid (Sigma-Aldrich). A combination of all three oligonucleotides containing the mutated or the wild-type binding sites was used in the assay. DNA-bound proteins were collected with streptavidin-agarose beads, washed with HKMG buffer, and finally resuspended in NuPAGE loading buffer (Invitrogen Life Technologies), heated to 70°C for 10 min, and separated on a NuPAGE 4–12% Bis-Tris gel (Invitrogen Life Technologies). The proteins were electroblotted onto a PVDF membrane (GE Healthcare) and detected using RUNX1 or RUNX3 antibodies described in the previous section.

Promoter enzyme immuno assay.
As performed in the pull-down assay, HEK293T cells were transfected with RUNX1 or RUNX3 and subsequently lysed. Insoluble material was removed by centrifugation. 384-well plates, precoated with streptavidin (Thermo Fisher Scientific) were washed 3 times with washing buffer (PBS and 0.05% Tween 20). Biotinylated FOXP3 promoter/oligonucleotides probes containing the RUNX binding sites were added (1 pmol per well; 50 fmol/µl) and incubated for 1 h at room temperature. Either a combination of all three oligonucleotides containing the mutated or the wild-type binding sites was used or single oligonucleotides were used in the assay. After 3 washing steps with washing buffer, the nuclear extract was added (concentration > 0.2 µg/µl) and incubated overnight at 4°C. The lysates were incubated with 10 µg of poly-deoxyinosinic-deoxycytidylic acid (Sigma-Aldrich). The plate was washed with HKMG buffer and incubated with a 1:1000 dilution of rabbit anti-RUNX1 (ab11903, Abcam) or 1:200 dilution of rabbit anti-RUNX3 (H-50, Santa Cruz Biotechnology, Inc.) at 4°C for 2 h. After three washing steps with HKMG buffer, a secondary antibody (anti–rabbit IgG-HRP, 1:3,000 in HKMG buffer, Cell Signaling Technology) was added, and the plate was incubated for 1 h at 4°C. The wells were washed 4 times with HKMG buffer before adding the substrate reagent (R&D Systems). The colorimetric reaction was stopped by adding 2 M H2SO4. Absorbance at 450 nm was measured using a microplate reader (Berthold Technologies).

ChIP.
Human naive CD4+ T cells were cultured either with IL-2 only or with IL-2, anti-CD2/3/28, and TGF-β for 72 h, and protein–DNA complexes were fixed by cross-linking with formaldehyde in a final concentration of 1.42% for 15 min. Formaldehyde was quenched with 125 mM glycine for 5 min, and cells were subsequently harvested. The ChIP assay was performed as described in the fast chromatin immunoprecipitation method (Nelson et al., 2006). Cells were lysed with immunoprecipitation buffer (150 mM NaCl, 50 mM Tris-HCl, pH 7.5, 5 mM EDTA, NP-40 [0.5% vol/vol]) containing phosphatase (Roche) and protease inhibitors cocktails (Roche), the nuclear pellet was washed, the chromatin was sheared by sonication and incubated with antibodies for RUNX1 (H-65 X; Santa Cruz Biotechnology, Inc.), RUNX3 (H-50 X; Santa Cruz Biotechnology, Inc.), CBFβ (PEBP2β; FL-182 X; Santa Cruz Biotechnology, Inc.), and as controls normal rabbit IgG (Santa Cruz Biotechnology, Inc.), anti-human RNA polymerase II antibody, and mouse control IgG (both from SA Biosciences). The cleared chromatin was incubated with protein A agarose beads and, after several washing steps, DNA was isolated with 10% (wt/vol) Chelex 100 resin. Samples were treated with proteinase K at 55°C for 30 min. The proteinase K was then inactivated by boiling the samples for 10 min. The purified DNA was used in a real-time PCR reaction. Specific primers for the FOXP3 promoter, spanning the region from −87 to −3, FOXP3 promoter forward primer 5′-AGAGGTCTGCGGCTTCCA-3′, FOXP3 promoter reverse primer 5′-GGAAACTGTCACGTATCAAAAACAA-3′, or control GAPDH primer (SA Biosciences) for the RNA polymerase II were used. A negative control PCR for each immunoprecipitation using IGX1A negative control primer targeting ORF-free intergenic DNA (SA Biosciences) was used. The fold enrichment in site occupancy was calculated incorporating IgG control values and input DNA values using the ChampionChIP qPCR data analysis file (SA Biosciences).

Quantification of cytokine levels.
IL-4, IL-5, IL-6, IL-10, IL-13, IL-17, and IFN-γ secretion was assessed using fluorescent bead-based technology. The Bio-Plex-hu Cytokine Panel, 17-Plex Group 1 was used according to the manufacturer's instructions (Bio-Rad Laboratories). Fluorescent signals were read and analyzed using the Bio-Plex 200 System (Bio-Rad Laboratories).

Online supplemental material.
Fig. S1 shows the induction of RUNX1, RUNX3, and FOXP3 mRNA in human CD4+ T cells after anti-CD2/3/28 mAb and TGF-β stimulation. Fig. S2 shows decreased RUNX1 and RUNX3 mRNA and protein expression after siRNA-mediated knockdown and decreased FOXP3 expression in human CD4+ T cells after RUNX1 and RUNX3 knockdown. Fig. S3 shows the quantification of IL-4, IL-5, IL-10, IL-13, and IFN-γ levels in control siRNA transfected or RUNX1 and RUNX3 siRNA transfected human CD4+ T cells. Fig. S4 shows the putative RUNX binding sites in the FOXP3 core promoter sequence of human, mouse, and rat. Fig. S5 shows the induction of FOXP3 protein after overexpression of RUNX1 and RUNX3 in human CD4+ T cells. Fig. S6 shows that endogenous IL-4 and IFN-γ do not effect Foxp3 expression in naive CD4+ T cells of CbfbF/F CD4-cre and CbfbF/F control mice, which were stimulated with anti-CD3 and anti-CD28 mAbs, IL-2 and TGF-β in the absence or presence of anti-IL-4 and anti-IFN-γ neutralizing mAbs. Fig. S7 shows similar cell death (A) and proliferation (B) of Foxp3+ and Foxp3− cells in CbfbF/F CD4-cre and CbfbF/F control mice cultures. Online supplemental material is available at http://www.jem.org/cgi/content/full/jem.20090596/DC1.