The FOXP3Δ2-expressing retrovirus (pRV-FOXP3Δ2-IRES-GFP) and the empty control virus (pRV-IRES-GFP) were generated as previously described [40, 41]. Briefly, the following primer pair was used to amplify the FOXP3Δ2 coding sequence from cDNA derived from human CD4+ T cells: 5′-cgggatccGGACAAGGACCCGATGCCCAACC-3′, 5′-CCCTGCCCCCACCACCTCTGC-3′. Retroviruses were transfected into PT67 packaging cell lines by calcium phosphate precipitation. Virus particles were produced by cultivating transfected PT67 cells for 24 h at 37°C in Iscove’s Modified Dulbecco’s Medium (IMDM) supplemented with 10% heat-inactivated fetal calf serum (FCS) and 100 U/ml penicillin/streptomycin. Remaining packaging cells were removed by filtering culture supernatants through a 0.2-μm syringe filter, after which 3 to 4 × 106 Jurkat T cells were resuspended in 2.5 ml of virus supernatant in a 6-well culture plate. Next, 2.5 μl polybrene (8 mg/ml Sigma-Aldrich, Steinheim, Germany) and 50 μl HEPES (Invitrogen, Darmstadt, Germany) were added per well. Cells were centrifuged at room temperature at 500 × g for 2 h and were further incubated for 24 h at 37°C. Transduction efficiency was checked by flow cytometry, and GFPhigh T cells were sorted with a BD FACSAria II cell sorter (Becton Dickinson [BD], Franklin Lakes, New Jersey, USA), followed by further cell expansion in IMDM.
Isolation of primary human T cells
CD4+CD25+FOXP3+ and naïve CD4+CD25+FOXP3- T cells were isolated from leukapheresis filters which were kindly provided by the Institute for Transfusion Medicine and Immune Hematology of the University Hospital Magdeburg. Cells were isolated using human CD4+CD25+ Regulatory T Cell Isolation Kit and AutoMACS device (both Miltenyi Biotec, Bergisch Gladbach, Germany) according to manufacturer’s instructions. Purity of isolated cells was checked by flow cytometry using CD3-FITC (clone OKT3, eBioscience, San Diego, CA, USA), CD4-PE-Cy7 (clone RPA-T4, eBioscience, San Diego, CA, USA) and CD25-PE (clone 4E3, Miltenyi Biotec, Bergisch Gladbach, Germany) antibody.
RNA was isolated by using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s recommendations. DNA was digested by using the RNase-Free DNase Set (Qiagen). RNA was eluted in 100 μl nuclease-free water and further concentrated by ethanol precipitation, beginning with the addition of 2 μl linear polyacrylamide (0.5 μg/μl,; Ambion, Darmstadt, Germany), 50 μl 7.5 M NH4OAc (Sigma-Aldrich), and 375 μl absolute ethanol, precooled to -20°C. RNA was precipitated for at least 2 h at -70°C and was then centrifuged at 14,000 × g at 4°C for 30 min in a table-top centrifuge. The RNA pellet was washed twice with 80% ethanol at -20°C, dried, and resuspended in nuclease-free water. RNA integrity was tested by using the Agilent 2100 Bioanalyzer (Agilent Technologies, Böblingen, Germany) with an Agilent RNA 6000 Nano/Pico Kit (Agilent Technologies).
Equal amounts of RNA were used for cDNA synthesis in a reverse transcription reaction using a mixture of 0.5 μl oligo-dT (0.5 μg/μl) and 0.5 μl random primers (3 μg/μl) and nuclease-free water (total volume, 12 μl). Samples were incubated for 10 min at 70°C and then put on ice for 10 min. Reverse transcription (RT) reaction mix was added, containing 4 μl 5x first-strand buffer (Invitrogen), 2 μl dithiothreitol (DTT; 0.1 M, Invitrogen), 1 μl dNTP Mix (10 mM, Invitrogen), and 1 μl SuperScript II Reverse Transcriptase (Invitrogen).
Samples were incubated for 60 min at 42°C in a thermocycler. Real-time PCR was performed in duplicate with LightCycler 480 SYBR Green I Master reaction mix (Roche, Mannheim, Germany) and a LightCycler 480 system (Roche). PCR reactions were performed with 1 μl cDNA, 5 μl primer mix (containing forward and reverse primer, 500 nmol/l each), 10 μl 2x concentrated LightCycler 480 SYBR Green I Master mix (Roche), and 4 μl nuclease-free water. Relative mRNA levels were determined by using standard curves for each gene, and quantitative normalization of gene expression was performed in relation to the expression of the housekeeping gene RPS9. As a template for the standard curves, a mixture of all samples in 4 serial dilutions was used, corresponding to 3, 1, 0.1, and 0.01 μl template amount. The following PCR temperature cycles were used (45 cycles were performed): Step 1: 95°C, 5 min; Step 2: 95°C, 10 sec; Step 3: 55-60°C (primer dependent), 10 sec; Step 4: 72°C, 10 sec; Step 5: 4°C, hold. Real-time PCR data were analyzed with LightCycler 480 software (Roche).
Confirmation of ChIP enrichment by site-specific genomic real-time PCR was performed by using amplified ChIP DNA from FOXP3-specific immunoprecipitation and from isotype control immunoprecipitation and in reference to an unrelated genomic region within the RPS9 gene. The following primers were used: RPS9 for: 5′-CGCAGGCGCAGACGGTGGAAGC-3′, RPS9 rev: 5′-CGAAGGGTCTCCGCGGGGTCACAT-3′, IL-2 for: 5′-GTCACAAACAGTGCACCTAC-3′, IL-2 rev: 5′-ATGGTTGCTGTCTCATCAGC-3′, FOXP3 for: 5′-GAACGCCATCCGCCACAACCTGA-3′, FOXP3 rev: 5′-CCCTGCCCCCACCACCTCTGC-3′, IL-22 for: 5′-CAACAGGCTAAGCACATGTCA-3′, IL-22 rev: 5′-ACTGTGTCCTTCAGCTTTTGC-3′, IL-26 for: 5′-AGCAACGATTCCAGAAGACC-3′, IL-26 rev: 5′-TGCAGTTGACCAAAAACGTC-3′, TGF-β2 for: 5′-CCAAAGGGTACAATGCCAAC-3′, TGF-β2 rev: 5′-CAGATGCTTCTGGATTTATGGTATT-3′, IL-22 chr12.119 for: 5′-AAGCCCACCTCCCAGGTCCC-3′, IL-22 chr12.119 rev: 5′-AGACAGCCAAAGCCTACTTCTGGT-3′
FOXP3 expression in transduced Jurkat T cells was detected by Western blot analysis. FOXP3 was detected by using monoclonal mouse anti-human FOXP3 antibody (clone 206D; Biolegend, San Diego, CA, USA; 1:1000 dilution in Tris buffered saline Tween 20 (TBS-T) buffer with 5% milk powder). For detection, a secondary polyclonal goat anti-mouse antibody conjugated to horseradish peroxidase (Dianova, Hamburg, Germany; 1:4000 in TBS-T buffer with 5% milk powder) was used.
Intracellular cytokine staining
For intracellular cytokine detection, cells were cultured in IMDM medium and re-stimulated with PMA (final concentration, 10 ng/ml) and ionomycin (final concentration, 1 μg/ml) for 4 h. For the final 2 h of incubation, brefeldin A (Sigma-Aldrich) was added to block the secretion of cytokines. Subsequently, cells were fixed with 1% (v/v) paraformaldehyde solution in phosphate-buffered saline (PBS). Cells were permeabilized with 0.1% IGEPAL (Sigma-Aldrich) in PBS for 5 min. IL-2 staining was performed for 30 min (mouse anti-human IL-2 APC, clone 5344.111; BD).
Jurkat T cells were cultivated in IMDM medium supplemented with 10% FCS (PAA Laboratories) and 100 U/ml penicillin/streptomycin. If stated, cells were stimulated with PMA (10 ng/ml, Sigma-Aldrich) and ionomycin (1 μg/ml, Sigma-Aldrich) for 4 h. Protein crosslinking was ensured by the addition of formaldehyde (1% v/v) and by incubation on a shaker for 10 min at room temperature. Formaldehyde was quenched for 5 min by the addition of glycine (125 mM final concentration). After centrifugation, the culture supernatant was removed, and cells were washed once with ice-cold PBS. Cells were lysed in IP cell-lysis Buffer (10 mM Tris HCL [pH 7.5], 10 mM NaCl, 3 mM MgCl2, 1 mM phenylmethanesulfonyl fluoride [PMSF] 0.5% v/v IGEPAL CA-630) and incubated on ice for 10 min. Lysed cells were centrifuged at 2500 rpm for 5 min at 4°C, and the supernatant was discarded. The chromatin material was obtained by lysing the cell nuclei with IP nuclei-lysis buffer (50 mM Tris HCL [pH 8.0], 10 mM EDTA, 1% [v/v] sodium dodecyl sulfate [SDS], complete protease inhibitors) and incubating them on ice for 10 min. Chromatin was diluted in IP dilution buffer (20 mM Tris HCl [pH 8], 2 mM EDTA, 1% v/v Triton X-100, 150 mM NaCl, 1 mM PMSF) and stored at -80°C.
ChIP chromatin corresponding to 5 × 107 cells was fragmented with a Branson Sonifier 250 (Branson Ultrasonics, Danbury, USA) equipped with a microtip. The following sonication conditions were applied: 60% duty cycle, maximum output power; 6 to 7 sonication cycles consisting of 30 sec on and 60 sec off. Average DNA fragment size was between 200 and 1000 bp. The IP process was performed essentially as described by the Affymetrix company: Protein G Sepharose 4B (Zymed, San Francisco, USA) and either 10 μg anti-human FOXP3 antibody (clone 206D, mouse IgG1,к; BioLegend) or 10 μg of isotype control antibody (clone MOPC-21, mouse IgG1,к; BioLegend) were used per IP, respectively.
After adapter ligation (adapter primer: 5′-GTTTCCCAGTCACGGTC(N)9-3′), ChIP DNA was amplified with a large-scale PCR reaction (amplification primer: 5′-GTTTCCCAGTCACGGTC-3′) involving Ampli Taq Gold DNA polymerase and partial incorporation of deoxyuridine triphosphate (dUTP). Amplified ChIP DNA was labelled and prepared for hybridization with the GeneChip WT Double-Stranded DNA Terminal Labeling Kit (Affymetrix) and was then analysed on a Human Promoter 1.0R tiling microarray (Affymetrix).
Tiling microarray data analysis
Basic tiling array data analysis was performed with TAS software (Version 1.1.02, Affymetrix). FOXP3-bound ChIP-enriched genomic regions were identified by comparing FOXP3 IP to a matched isotype control IP from the same cell material. Only continuous genomic regions exhibiting a p-value higher than the 95th percentile of all p-values were included in subsequent analyses. Both resting and PMA/ionomycin–stimulated conditions were analyzed in independent duplicate experiments. Each replicate experiment was analyzed separately. Calculated ChIP intervals were visualized with IGB software . The web-based Galaxy analysis tool  was used to compare repetitive ChIP-on-chip experiments. Positional intersecting ChIP regions between two replicate experiments were calculated; thus, these regions represented genomic regions that were identified in both ChIP-on-chip replicates. Moreover, the range of intersection was calculated, and only the corresponding genomic coordinates of intersecting areas were considered for further analysis. The tiling microarray data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus [44, 45] and are accessible through GEO Series accession number GSE37256 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE37256). Significance of the overlap between different lists of HGNC gene symbols was calculated using a hypergeometric test.
Expression microarray analysis
Equal amounts of RNA were amplified with a double linear amplification protocol, starting with the synthesis of double-stranded cDNA by using the T7dT23 primer (5′-GGCCAGTGAATTGTAATACGACTCACTATAGGGAGGCGG(T)23-3′; Metabion, Planegg, Germany) and SuperScript II reverse transcriptase (Invitrogen, Karlsruhe, Germany). A second-strand cDNA synthesis was then performed with DNA polymerase I (Invitrogen) and E. coli DNA ligase (Invitrogen). After purification of double-stranded cDNA, the first amplification round was performed in an in vitro transcription reaction by using the Promega P1300 RiboMax Kit for T7 amplification (Promega, Mannheim, Germany); this procedure produced unlabeled cRNA. The amplified cRNA was subjected to a second amplification round, starting again with reverse transcription by using random hexamer primers (Pharmacia, Freiburg, Germany) for first-strand synthesis of cDNA, followed by cRNA degradation with RNase H treatment. T7dT23 primers were again used for the second-strand synthesis, as described above. In the second in vitro transcription, which used the GeneChip expression 3′-Amplification Reagent Kit for labeling (Affymetrix, San Francisco, CA, USA), biotinylated UTP was partially incorporated into the final cRNA. The quantity and quality of biotinylated cRNA were checked photometrically. Biotinylated cRNA was fragmented, hybridized to a Human Genome U133 Plus 2.0 Array (Affymetrix), washed, and stained as recommended by the manufacturer. Microarray data were analyzed with GeneSpring GX 10.0 software (Agilent Technologies) and the Robust Multi-array Analysis (RMA) normalization algorithm. Cluster analysis was performed with Genesis Software 1.7.3  by using a z-score transformation . A list of 410 FOXP3-dependent genes was compiled by combining two analysis strategies. First, J-FOXP3 cells were compared with J-GFP cells. Genes were considered to be influenced by the over-expressed FOXP3 protein only if their expression showed at least a two-fold change (up or down) under either stimulated or unstimulated conditions. In a second strategy, stimulated and resting J-FOXP3 and stimulated and resting J-GFP cells were compared by calculating the differences in gene expression. Only genes that showed at least a two-fold change in stimulation-dependent induction or repression were considered to be influenced by the over-expressed protein. Subsequently this list of stimulation-dependent transcripts was filtered for differences between FOXP3-expressing cells and GFP control cells. Only transcripts that exhibited at least a two-fold difference in stimulation dependency between J-FOXP3 and J-GFP cells were retained. Both analysis strategies resulted in a combined list of 410 FOXP3-dependent genes. The expression microarray data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus [44, 45] and are accessible through GEO Series accession number GSE37256 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE37256).