PMC:2806624 / 5858-26996 JSON TXT

Annnotations TAB JSON ListView MergeView

2_test

{"project":"2_test","denotations":[{"id":"19917773-16387603-59851498","span":{"begin":5031,"end":5035},"obj":"16387603"},{"id":"19917773-17785809-59851499","span":{"begin":13524,"end":13528},"obj":"17785809"},{"id":"19917773-16648838-59851500","span":{"begin":16890,"end":16894},"obj":"16648838"}],"text":"RESULTS\n\nThe role of RUNX1 and RUNX3 transcription factors in TGF-β–mediated iT reg cell generation\nTo investigate the role of RUNX transcription factors in the development of iT reg cells, we cultured naive CD4+ T cells, isolated from human PBMCs, in conditions that enable the development of iT reg cells. Stimulation with anti-CD2/3/28 mAbs or TGF-β alone resulted in a minimal up-regulation of RUNX1 and RUNX3 mRNA (Fig. 1 A). In contrast, the combination of both TGF-β and anti-CD2/3/28 mAbs induced RUNX1 and RUNX3 mRNAs, as well as FOXP3 mRNA, within 48 h in naive CD4+ T cells. This result suggested further experiments to investigate whether the up-regulation of RUNX1 and RUNX3 might be a feature of iT reg cells during their development or even a prerequisite for their induction.\nFigure 1. RUNX1 and RUNX3 are involved in the induction of Foxp3 in iT reg cells. (A) RUNX1, RUNX3, and FOXP3 mRNA induction in human naive CD4+ T cells after alone or combined anti-CD2/3/28 mAb and TGF-β stimulation in the presence of IL-2. Real-time PCR of human naive CD4+ T cells after 48 h of culture. Bars show the mean ± SE of three independent experiments. (B) FOXP3 mRNA induction by anti-CD2/3/28 mAb and TGF-β in human naive CD4+ T cells is reduced after siRNA-mediated RUNX1/3 knockdown. Real-time PCR of RNA from human naive CD4+ T cells, transfected with RUNX1 and/or RUNX3 siRNA or with a control siRNA and cultured with anti-CD2/3/28, TGF-β and IL-2. Bars show the mean ± SD of three independent experiments. (C) FOXP3 mRNA is down-regulated in iT reg cells after siRNA-mediated knockdown of RUNX1 and RUNX3. Real-time PCR for FOXP3, T-bet, GATA3, and RORC2 from human naive CD4+ T cells, transfected with RUNX1 and RUNX3 siRNA or with scrambled siRNA (control) and cultured under iT reg, Th1, Th2, or Th17-driving conditions for 12 d. Bars show the mean ± SD of three independent experiments. (D) FOXP3 protein induction in iT reg cells is reduced after siRNA-mediated RUNX1/3 knockdown. Human naive CD4+ T cells were transfected with RUNX1 and/or RUNX3 siRNA or with a control siRNA and stimulated with anti-CD2/3/28 and TGF-β in the presence of IL-2. CD4 and intracellular FOXP3 analysis by flow cytometry after 72 h. One of three independent experiments is shown. Statistical differences were verified by the paired Student's t test. *, P \u003c 0.05; **, P \u003c 0.01. To test this hypothesis, RUNX1 and RUNX3 expression was knocked down in human naive CD4+ T cells by transfection of small interfering RNAs (siRNAs; Fig. 1 B). Deficiency of RUNX1 or RUNX3 resulted in markedly reduced TGF-β–mediated induction of FOXP3 mRNA in naive CD4+ T cells compared with control cells transfected with scrambled siRNA. The level of FOXP3 mRNA was further reduced when both RUNX1 and RUNX3 were knocked down in naive CD4+ T cells during their differentiation to iT reg cells (Fig. 1 B).\nThe influence of RUNX1 and RUNX3 on the development of other T cell subsets and their specific transcription factor expression was further investigated. Naive CD4+ T cells were cultured under Th1, Th2, T reg cell, and Th17 differentiation conditions and the mRNA expression of the predominant transcription factor for each cell type was subsequently analyzed. We observed no change in GATA3 expression in Th2 cells, T-bet expression in Th1 cells, or RORC2 mRNA expression in Th17 cells in which RUNX1 and RUNX3 were knocked down compared with control cells. On the contrary, FOXP3 mRNA was significantly decreased in RUNX1- and RUNX3-deficient T reg cells compared with control cells (Fig. 1 C).\nThe effect of RUNX silencing on the expression level of intracellular FOXP3 during naive CD4+ T cell differentiation to iT reg was evaluated by flow cytometry. FOXP3 was only slightly reduced after RUNX1 silencing. Transfection of siRNA for RUNX3 had a stronger effect. The most striking FOXP3 reduction was observed when RUNX1 and RUNX3 were silenced together (Fig. 1 D). Similar results were obtained in total CD4+ T cells (Fig. S1 and Fig. S2). The increased impact of combined RUNX1 and RUNX3 knockdown implies that RUNX1 and RUNX3 might have redundant functions in the induction of FOXP3. In addition, the levels of IL-4, IL-5, IL-10, IL-13, and IFN-γ in control siRNA-transfected or RUNX1 and RUNX3 siRNA-transfected CD4+ T cells that were cultured with or without anti-CD2/3/28 mAb and TGF-β did not show any significant difference (Fig. S3).\nTo determine whether RUNX1 and RUNX3 are also expressed in human T reg cells in vivo, we isolated peripheral blood CD4+ CD127− CD25high T reg cells and compared them with CD4+ CD127+ CD25− T cells. Circulating T reg cells expressed significantly higher levels of RUNX3 mRNA compared with CD4+CD25− cells. As expected, IL-10, TGF-β, and FOXP3 mRNAs are also expressed in circulating T reg cells (Fig. 2 A). There was no difference in RUNX1 mRNA expression between these two cell subsets. We also performed an analysis of human tonsils, which contain high numbers of FOXP3+ T reg cells (Verhagen et al., 2006). Staining of tonsil sections for FOXP3 and RUNX3 demonstrated in vivo coexpression of these two molecules in a subset of T reg cells, whereas there was low RUNX1 expression in all tonsil cells (Fig. 2 B).\nFigure 2. RUNX1 and RUNX3 expression in Foxp3+ T reg cells and in human CD4+, CD127−, CD25high cells. (A) Real-time PCR analysis of CD4+, CD127−, and CD25high T reg cells and CD4+, CD127+, and CD25neg T cells isolated from human peripheral blood showed an increased expression of IL-10, TGF-β, FOXP3, and RUNX3 mRNA in CD25+ compared with CD25− cells. Bars show the mean ± SE of three independent experiments. (B) Human tonsil sections were analyzed by confocal microscopy. Tissue sections were stained for FOXP3, RUNX1, RUNX3, and DAPI or isotype controls. HEK cells RUNX1-transfected or not transfected served as additional control for RUNX1 staining. Data shown are representative from one of the three tissue samples with similar results. Bars, 5 µm. Statistical differences were verified by the paired Student's t test. *, P \u003c 0.05; **, P \u003c 0.01.\n\nRUNX1 and RUNX3 bind to the FOXP3 promoter\nTranscription element search system analysis of the human FOXP3 promoter predicted 3 putative RUNX binding sites at 333, 287, and 53 bp upstream of the transcription start site (TSS). All three binding sites are conserved between human, mouse, and rat (Fig. S4). To verify the putative binding sites in the FOXP3 promoter, we transiently transfected HEK293T cells with RUNX1 and RUNX3. After the pull-down with oligonucleotides containing the wild-type binding sequences, but not mutant sequences, RUNX binding to the FOXP3 promoter oligonucleotides was detected by Western blot (Fig. 3 A). To confirm these results and test the ability of single binding site sequences to bind either RUNX1 or RUNX3, we used the promoter enzyme immunoassay. Cell lysates were obtained from HEK293T cells that had been transiently transfected with RUNX1 or RUNX3 expression vectors. The biotinylated FOXP3 promoter oligonucleotides were linked to a streptavidin-coated microtiter plate, and bound RUNX1 or RUNX3 was detected by using anti-RUNX antibodies and a peroxidase-labeled secondary antibody. We showed binding of RUNX1 and RUNX3 to the mixture of all three oligonucleotides containing the binding sites, whereas there was no binding detectable when a combination of the mutated oligonucleotides was used in the assay (Fig. 3 B). Although there was a similar and high degree of binding to the −333 and −287 sites, a lower degree of binding was detected when the oligonucleotide containing the −53 site in the Foxp3 promoter was used. This effect was observed both for binding to RUNX1 and RUNX3 (Fig. 3 B). The binding to the two single binding sites at −333 and −287 was comparable to the mixture of all three oligonucleotides. Chromatin immunoprecipitation (ChIP) assay results confirmed the binding of RUNX1 and RUNX3 complexes containing CBFβ to FOXP3 promoter during the differentiation of naive T cells toward T reg cells. Here, naive CD4+ T cells were cultured with IL-2, anti-CD2/3/28 mAb, and TGF-β as a Foxp3-inducing stimulation. Amplification of PCR products from the FOXP3 promoter region with the predicted RUNX binding sites showed that RUNX1, RUNX3, and CBFβ were immunoprecipitated together with the FOXP3 promoter (Fig. 3 C). Negative control primer targeting open reading frame-free intergenic DNA, IGX1A did not show any significant change in site occupancy.\nFigure 3. Binding of RUNX1 and RUNX3 proteins to the predicted binding sites in the FOXP3 promoter. (A) Mutated and wild-type oligonucleotides are shown. The predicted RUNX binding sites are accentuated (boxed and in green letters) and stars mark mutations introduced into the binding site of the control oligonucleotides. Nuclear extracts from HEK293T cells were incubated with biotinylated oligonucleotides. The precipitated oligonucleotide–transcription factor complexes were separated by SDS-PAGE and identified by Western blotting with anti-RUNX1 and anti-RUNX3 antibodies. A mixture of all three oligonucleotides with the predicted binding sites or with the inserted mutation into the predicted sites was used. Data shown are one representative of three independent experiments with similar results. (B) Promoter enzyme immunoassay using wild-type and mutated oligonucleotides within the FOXP3 promoter. Bars show mean ± SE of three independent experiments. (C) Chromatin immunoprecipitation assay results show binding of RUNX1 and RUNX3 complexes containing CBFβ to the human FOXP3 promoter in naive CD4+ T cells that were cultured with IL-2 together with anti-CD2/3/28 and TGF-β. There was no change in site occupancy in all immunoprecipitations when IGX1A negative control primers were used. The results are normalized to input and isotype control antibody. Bars show mean ± SE of three independent experiments. Statistical differences were verified by the paired Student's t test. *, P \u003c 0.05\n\nRegulation of FOXP3 promoter activity and FOXP3 protein expression by RUNX1 and RUNX3\nTo investigate the effect of RUNX1 and RUNX3 binding to the RUNX binding sites in the FOXP3 promoter, we transfected human peripheral blood CD4+ T cells with a FOXP3 promoter luciferase reporter vector and RUNX1 or RUNX3 expression vectors. An increase in luciferase activity was observed only when the FOXP3 promoter (−511 to +176) luciferase construct was cotransfected with RUNX1 or RUNX3 expression vectors (Fig. 4 A). The increase in promoter activity was greater upon cotransfection of RUNX3 compared with RUNX1. Luciferase expression was abrogated when the Runx binding sites in the FOXP3 promoter (−511 to +176) luciferase construct were mutated (Fig. 4 A). In these experiments, the overexpression of RUNX1 and RUNX3 eliminated the need of TGF-β for FOXP3 promoter activation and PMA/ionomycin stimulation was sufficient.\nFigure 4. Regulation of FOXP3 promoter activity by RUNX1 and RUNX3. (A) Human primary CD4+ cells were transfected with an empty vector (pGL3 Basic), a vector containing the wild-type or mutated FOXP3 promoter region (FOXP3 −511/+176) fused to the luciferase reporter gene together with a GFP, RUNX1, or RUNX3 expression vector. Bars show the mean luciferase activity ± SE measured as arbitrary light units of three independent experiments. (B) Human primary CD4+ cells were transfected with an empty vector (pGL3 Basic), a vector containing the putative FOXP3 promoter region (FOXP3 −511/+176) fused to the luciferase reporter gene, or with a vector containing the putative FOXP3 promoter region (FOXP3 −511/+176) with single RUNX binding sites mutated (53, 287, or 333) or with the combination of two or three RUNX binding sites mutated (53, 287, or 333) fused to the luciferase reporter gene. Bars show the mean luciferase activity ± SD measured as arbitrary light units of three independent experiments. To examine the role of each of the three RUNX binding sites for the FOXP3 promoter activity, we mutated each individually or in combination. No reduction in luciferase activity was observed when the −53 site was mutated and only a slight reduction when either the −287 or −333 site was mutated (Fig. 4 B). However, mutating the −53 site in combination with one of the other two sites led to a significant decrease in luciferase activity, with the greatest reduction observed when all three binding sites were mutated (Fig. 4 B), suggesting that the identified binding sites have redundant functions and RUNX binding to more than one site is necessary for the full activation of the FOXP3 promoter. Supporting these findings, the overexpression of RUNX1 in human primary CD4+ T cells resulted in significantly elevated levels of FOXP3 protein measured by flow cytometry after 48 h. This was achieved without any requirement for anti-CD3, anti-CD28 stimulation, or the presence of TGF-β. Although there was a trend, the transfection of CD4+ T cells with RUNX3 did not lead to statistically significant increase in FOXP3 (Fig. S5).\n\nRole of CBFβ in the induction of Foxp3\nCBFβ, a common cofactor of all RUNX proteins, stabilizes and increases the binding of the runt domain to target DNA sequences. To target all Runx proteins that might be involved in the induction of Foxp3, we used mice in which loxP-flanked Cbfb alleles were inactivated in T cells through expression of a CD4-cre transgene. Retinoic acid and TGF-β synergize in the induction of Foxp3 in naive T cells (Kang et al., 2007). To investigate whether Runx-mediated induction of Foxp3 is dependent on the expression of CBFβ, naive CD4+ CD8− T cells from CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice were stimulated with anti-CD3/28 mAbs, retinoic acid, and increasing concentrations of TGF-β. After 3 d in culture, the cells were restimulated with PMA and ionomycin and analyzed for intracellular Foxp3 and IFN-γ expression. TGF-β induced Foxp3 in CbfbF/+ CD4-cre cells in a dose dependent manner, and this was significantly reduced in CbfbF/F CD4-cre cells. Retinoic acid enhanced Foxp3 expression even in 20 pg/ml of TGF-β and more than 95% of the CD4+ T cells from CbfbF/+ CD4-cre mice became Foxp3+ in 100 and 500 pg/ml TGF-β doses. The induction of Foxp3 was again significantly lower in CbfbF/F CD4-cre CD4+ T cells even in the presence of retinoic acid, demonstrating that deficiency in Runx binding to DNA affects the TGF-β induction of Foxp3 in T reg cells (Fig. 5 A). There was no difference in the induction of Foxp3 when endogenous IL-4 and IFN-γ were neutralized (Fig. S6).\nFigure 5. Diminished capacity of Cbfb-deficient CD4-cre mice T cells in the generation of Foxp3+ CD4+ T cells. (A) FACS-purified naive CD4+ CD8− T cells from CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice were activated in vitro with anti-CD3/28 mAb, 50 U/ml IL-2, ± 10 nM retinoic acid (RA), and increasing concentrations of TGF-β. After 3 d in culture, the cells were restimulated with PMA + ionomycin, and then analyzed for intracellular Foxp3 and IFN-γ expression. One of five experiments is shown. (B) 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 per transfer). 6 wk later, TCRβ+CD4+ cells from the spleen, mesenteric lymph node (MLN), and lamina propria of the small intestine (LP) were analyzed for Foxp3-GFP expression. Results from one of four CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice with same findings are shown. The data from four sets of mice is shown in C. Statistical analysis was performed with Mann-Whitney U test. *, P \u003c 0.05 between groups. The importance of RUNX transcription factors for the in vivo conversion of naive CD4+ T cells into iT reg cells was examined. Control CbfbF/F or CbfbF/F CD4-cre naive T cells, harboring a Foxp3-IRES-GFP allele were adoptively transferred into Rag2−/− mice. 6 wk later, CD4+ T cells in spleen, mesenteric lymph node, and lamina propria of the small intestine were analyzed for Foxp3-GFP expression (Fig. 5 B). There was a consistently lower percentage of CD4+ T cells that had developed Foxp3 expression in the mesenteric lymph node and lamina propria of mice transferred with CbfbF/F CD4-cre cells compared with control cells (Fig. 5, B and C). These data affirm the significance of RUNX proteins for the in vivo generation of CD4+ Foxp3+ T cells.\n\nCBFβ, RUNX1, and RUNX3 are important for the suppressive activity of CD25+ Foxp3+ T reg cells\nEven though Cbfb-deficient CD4+ T cells had impaired induction of Foxp3 after stimulation with anti-CD3/28 mAbs and TGF-β, sufficient numbers of Foxp3+ cells could be generated to permit analysis of their suppressive activity. Purified naive CD4+ T cells from CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice (Cd45.2) harboring a Foxp3-ires-GFP allele (Bettelli et al., 2006) were stimulated in vitro with anti-CD3/28 mAbs, IL-2, and TGF-β. After 3 d, Foxp3+-GFP+ cells were sorted by flow cytometry and mixed with CFSE-labeled naive CD45.1+ CD4+ cells at ratios of 1:4, 1:2, and 1:1. The cells were then incubated with inactivated splenocytes and stimulated with anti-CD3 mAb for four more days. CD45.1+ cells were analyzed for CFSE dilution (Fig. 6 A). CbfbF/+ CD4-cre CD4+ T cells activated in the presence of TGF-β showed a clear suppression of T cell proliferation that became even more apparent when an increased ratio of FOXP3+/CD25− cells was used (Fig. 6 B). The suppression was significantly reduced when cells from CbfbF/F CD4-cre mice were used, demonstrating that TGF-β–induced Runx complexes are important for the suppressive activity of Foxp3+ T reg cells (Fig. 6 B). As a control, CbfbF/+ CD4-cre CD4+ T cells activated in absence of TGF-β were mixed with CFSE-labeled naive CD45.1+ CD4+ cells. No suppression could be observed in all control groups without TGF-β at all tested ratios (Fig. 6 C). The decreased suppression capacity of CbfbF/F CD4-cre iT reg cells was unlikely to be caused by decreased survival or proliferation. Foxp3+ and Foxp3− cells generated from both CbfbF/F CD4-cre and control cells all displayed similar proliferation rates and cell death as measured by CFSE dilution and annexin V staining, respectively (Fig. S7, A and B).\nFigure 6. CbfbF/F CD4-cre mouse cells and iT reg cells generated from human naive CD4+ T cells undergoing siRNA-mediated RUNX1 and RUNX3 knock down show a diminished suppressive activity. Experimental setup (A) and results of the mouse suppression assay (B), FACS-purified naive CD4+8− T cells from CbfbF/F CD4-cre (left) and control CbfbF/+ CD4-cre mice (Cd45.2; right) were activated in vitro with anti-CD3/28 mAb, 50 U/ml IL-2, and 2.5 ng/ml TGF-β. After 3 d, Foxp3-GFP+ cells were FACS-sorted and mixed with CFSE-loaded naive CD45.1+ CD4+ cells at the indicated ratios. These were then incubated with inactivated splenocytes and anti-CD3 mAb. After a further 4 d, CD45.1+ cells were analyzed for CFSE dilution. (C) As a control, CbfbF/+ CD4-cre CD4+ T cells activated in absence of TGF-β were mixed with CFSE-loaded naive CD45.1+ CD4+ cells at the indicated ratios. Four days later CD45.1+ cells were analyzed for CFSE dilution. The median division number (of the naive CD45.1+ CD4+ cells in the cultures containing Foxp3-GFP+ cells) is indicated in each of the histograms. One of three experiments is shown. (D) Human naive CD4+ T cells, transfected with RUNX1 and RUNX3 siRNA and cultured under iT reg differentiating conditions were used in an in vitro suppression assay, cultured together with autologous CFSE-labeled CD4+ T cells, and stimulated with anti-CD3 mAb. The CFSE dilution of the CD4+ T cell responder cells was analyzed after 5 d by flow cytometry. The T reg/responder CD4+ T cell ratios used were 1:20, 1:10, and 1:5. One of two experiments is shown. In addition, we tested the requirement of RUNX1 and RUNX3 for the development of the suppressive capacity in human iT reg cells. We isolated human naive CD4+ T cells and transfected them with a combination of RUNX1 and RUNX3 siRNA, or with a scrambled control siRNA. Cells were cultured under T reg conditions, and then mixed with CFSE-labeled autologous CD4+ T cells and stimulated with anti-CD3 mAb. Cells in which RUNX1 and RUNX3 were knocked down showed markedly lower suppressive activity compared with control iT reg cells at a T reg/CD4+ T responder cell ratio of 1:20, but not when the T reg/CD4+ T responder cell ratio was increased to 1:5 (Fig. 6 D). These results demonstrate the important role of RUNX1 and RUNX3 not only for the induction of FOXP3, but also for the suppressive capacity of iT reg cells both in humans and in mice. The data suggest both quantity and quality of T reg cells are hampered. Reduced intrinsic suppressive capacity of iT reg cells was demonstrated in mice, because Foxp3-GFP+ cells were FACS sorted and same numbers of iT reg cells are included in control experiments. In the suppression experiment with human cells, reduced suppressive activity was caused by reduced FOXP3 expression in T cells by siRNA inhibition of RUNX1 and RUNX3.\n"}

pmc-enju-pas

bionlp-st-ge-2016-coref

{"project":"bionlp-st-ge-2016-coref","denotations":[{"id":"T7915","span":{"begin":12821,"end":12825},"obj":"Anaphor"},{"id":"T7916","span":{"begin":12749,"end":12757},"obj":"Antecedent"},{"id":"T9452","span":{"begin":14008,"end":14012},"obj":"Anaphor"},{"id":"T9453","span":{"begin":13937,"end":13944},"obj":"Antecedent"},{"id":"T4457","span":{"begin":5123,"end":5142},"obj":"Anaphor"},{"id":"T4458","span":{"begin":5070,"end":5075},"obj":"Antecedent"},{"id":"T4459","span":{"begin":5080,"end":5085},"obj":"Antecedent"}],"relations":[{"id":"R6397","pred":"boundBy","subj":"T7915","obj":"T7916"},{"id":"R7550","pred":"boundBy","subj":"T9452","obj":"T9453"},{"id":"R3567","pred":"boundBy","subj":"T4457","obj":"T4458"},{"id":"R3568","pred":"boundBy","subj":"T4457","obj":"T4459"}],"namespaces":[{"prefix":"_base","uri":"https://bionlp.dbcls.jp/ontology/ge.owl#"}],"text":"RESULTS\n\nThe role of RUNX1 and RUNX3 transcription factors in TGF-β–mediated iT reg cell generation\nTo investigate the role of RUNX transcription factors in the development of iT reg cells, we cultured naive CD4+ T cells, isolated from human PBMCs, in conditions that enable the development of iT reg cells. Stimulation with anti-CD2/3/28 mAbs or TGF-β alone resulted in a minimal up-regulation of RUNX1 and RUNX3 mRNA (Fig. 1 A). In contrast, the combination of both TGF-β and anti-CD2/3/28 mAbs induced RUNX1 and RUNX3 mRNAs, as well as FOXP3 mRNA, within 48 h in naive CD4+ T cells. This result suggested further experiments to investigate whether the up-regulation of RUNX1 and RUNX3 might be a feature of iT reg cells during their development or even a prerequisite for their induction.\nFigure 1. RUNX1 and RUNX3 are involved in the induction of Foxp3 in iT reg cells. (A) RUNX1, RUNX3, and FOXP3 mRNA induction in human naive CD4+ T cells after alone or combined anti-CD2/3/28 mAb and TGF-β stimulation in the presence of IL-2. Real-time PCR of human naive CD4+ T cells after 48 h of culture. Bars show the mean ± SE of three independent experiments. (B) FOXP3 mRNA induction by anti-CD2/3/28 mAb and TGF-β in human naive CD4+ T cells is reduced after siRNA-mediated RUNX1/3 knockdown. Real-time PCR of RNA from human naive CD4+ T cells, transfected with RUNX1 and/or RUNX3 siRNA or with a control siRNA and cultured with anti-CD2/3/28, TGF-β and IL-2. Bars show the mean ± SD of three independent experiments. (C) FOXP3 mRNA is down-regulated in iT reg cells after siRNA-mediated knockdown of RUNX1 and RUNX3. Real-time PCR for FOXP3, T-bet, GATA3, and RORC2 from human naive CD4+ T cells, transfected with RUNX1 and RUNX3 siRNA or with scrambled siRNA (control) and cultured under iT reg, Th1, Th2, or Th17-driving conditions for 12 d. Bars show the mean ± SD of three independent experiments. (D) FOXP3 protein induction in iT reg cells is reduced after siRNA-mediated RUNX1/3 knockdown. Human naive CD4+ T cells were transfected with RUNX1 and/or RUNX3 siRNA or with a control siRNA and stimulated with anti-CD2/3/28 and TGF-β in the presence of IL-2. CD4 and intracellular FOXP3 analysis by flow cytometry after 72 h. One of three independent experiments is shown. Statistical differences were verified by the paired Student's t test. *, P \u003c 0.05; **, P \u003c 0.01. To test this hypothesis, RUNX1 and RUNX3 expression was knocked down in human naive CD4+ T cells by transfection of small interfering RNAs (siRNAs; Fig. 1 B). Deficiency of RUNX1 or RUNX3 resulted in markedly reduced TGF-β–mediated induction of FOXP3 mRNA in naive CD4+ T cells compared with control cells transfected with scrambled siRNA. The level of FOXP3 mRNA was further reduced when both RUNX1 and RUNX3 were knocked down in naive CD4+ T cells during their differentiation to iT reg cells (Fig. 1 B).\nThe influence of RUNX1 and RUNX3 on the development of other T cell subsets and their specific transcription factor expression was further investigated. Naive CD4+ T cells were cultured under Th1, Th2, T reg cell, and Th17 differentiation conditions and the mRNA expression of the predominant transcription factor for each cell type was subsequently analyzed. We observed no change in GATA3 expression in Th2 cells, T-bet expression in Th1 cells, or RORC2 mRNA expression in Th17 cells in which RUNX1 and RUNX3 were knocked down compared with control cells. On the contrary, FOXP3 mRNA was significantly decreased in RUNX1- and RUNX3-deficient T reg cells compared with control cells (Fig. 1 C).\nThe effect of RUNX silencing on the expression level of intracellular FOXP3 during naive CD4+ T cell differentiation to iT reg was evaluated by flow cytometry. FOXP3 was only slightly reduced after RUNX1 silencing. Transfection of siRNA for RUNX3 had a stronger effect. The most striking FOXP3 reduction was observed when RUNX1 and RUNX3 were silenced together (Fig. 1 D). Similar results were obtained in total CD4+ T cells (Fig. S1 and Fig. S2). The increased impact of combined RUNX1 and RUNX3 knockdown implies that RUNX1 and RUNX3 might have redundant functions in the induction of FOXP3. In addition, the levels of IL-4, IL-5, IL-10, IL-13, and IFN-γ in control siRNA-transfected or RUNX1 and RUNX3 siRNA-transfected CD4+ T cells that were cultured with or without anti-CD2/3/28 mAb and TGF-β did not show any significant difference (Fig. S3).\nTo determine whether RUNX1 and RUNX3 are also expressed in human T reg cells in vivo, we isolated peripheral blood CD4+ CD127− CD25high T reg cells and compared them with CD4+ CD127+ CD25− T cells. Circulating T reg cells expressed significantly higher levels of RUNX3 mRNA compared with CD4+CD25− cells. As expected, IL-10, TGF-β, and FOXP3 mRNAs are also expressed in circulating T reg cells (Fig. 2 A). There was no difference in RUNX1 mRNA expression between these two cell subsets. We also performed an analysis of human tonsils, which contain high numbers of FOXP3+ T reg cells (Verhagen et al., 2006). Staining of tonsil sections for FOXP3 and RUNX3 demonstrated in vivo coexpression of these two molecules in a subset of T reg cells, whereas there was low RUNX1 expression in all tonsil cells (Fig. 2 B).\nFigure 2. RUNX1 and RUNX3 expression in Foxp3+ T reg cells and in human CD4+, CD127−, CD25high cells. (A) Real-time PCR analysis of CD4+, CD127−, and CD25high T reg cells and CD4+, CD127+, and CD25neg T cells isolated from human peripheral blood showed an increased expression of IL-10, TGF-β, FOXP3, and RUNX3 mRNA in CD25+ compared with CD25− cells. Bars show the mean ± SE of three independent experiments. (B) Human tonsil sections were analyzed by confocal microscopy. Tissue sections were stained for FOXP3, RUNX1, RUNX3, and DAPI or isotype controls. HEK cells RUNX1-transfected or not transfected served as additional control for RUNX1 staining. Data shown are representative from one of the three tissue samples with similar results. Bars, 5 µm. Statistical differences were verified by the paired Student's t test. *, P \u003c 0.05; **, P \u003c 0.01.\n\nRUNX1 and RUNX3 bind to the FOXP3 promoter\nTranscription element search system analysis of the human FOXP3 promoter predicted 3 putative RUNX binding sites at 333, 287, and 53 bp upstream of the transcription start site (TSS). All three binding sites are conserved between human, mouse, and rat (Fig. S4). To verify the putative binding sites in the FOXP3 promoter, we transiently transfected HEK293T cells with RUNX1 and RUNX3. After the pull-down with oligonucleotides containing the wild-type binding sequences, but not mutant sequences, RUNX binding to the FOXP3 promoter oligonucleotides was detected by Western blot (Fig. 3 A). To confirm these results and test the ability of single binding site sequences to bind either RUNX1 or RUNX3, we used the promoter enzyme immunoassay. Cell lysates were obtained from HEK293T cells that had been transiently transfected with RUNX1 or RUNX3 expression vectors. The biotinylated FOXP3 promoter oligonucleotides were linked to a streptavidin-coated microtiter plate, and bound RUNX1 or RUNX3 was detected by using anti-RUNX antibodies and a peroxidase-labeled secondary antibody. We showed binding of RUNX1 and RUNX3 to the mixture of all three oligonucleotides containing the binding sites, whereas there was no binding detectable when a combination of the mutated oligonucleotides was used in the assay (Fig. 3 B). Although there was a similar and high degree of binding to the −333 and −287 sites, a lower degree of binding was detected when the oligonucleotide containing the −53 site in the Foxp3 promoter was used. This effect was observed both for binding to RUNX1 and RUNX3 (Fig. 3 B). The binding to the two single binding sites at −333 and −287 was comparable to the mixture of all three oligonucleotides. Chromatin immunoprecipitation (ChIP) assay results confirmed the binding of RUNX1 and RUNX3 complexes containing CBFβ to FOXP3 promoter during the differentiation of naive T cells toward T reg cells. Here, naive CD4+ T cells were cultured with IL-2, anti-CD2/3/28 mAb, and TGF-β as a Foxp3-inducing stimulation. Amplification of PCR products from the FOXP3 promoter region with the predicted RUNX binding sites showed that RUNX1, RUNX3, and CBFβ were immunoprecipitated together with the FOXP3 promoter (Fig. 3 C). Negative control primer targeting open reading frame-free intergenic DNA, IGX1A did not show any significant change in site occupancy.\nFigure 3. Binding of RUNX1 and RUNX3 proteins to the predicted binding sites in the FOXP3 promoter. (A) Mutated and wild-type oligonucleotides are shown. The predicted RUNX binding sites are accentuated (boxed and in green letters) and stars mark mutations introduced into the binding site of the control oligonucleotides. Nuclear extracts from HEK293T cells were incubated with biotinylated oligonucleotides. The precipitated oligonucleotide–transcription factor complexes were separated by SDS-PAGE and identified by Western blotting with anti-RUNX1 and anti-RUNX3 antibodies. A mixture of all three oligonucleotides with the predicted binding sites or with the inserted mutation into the predicted sites was used. Data shown are one representative of three independent experiments with similar results. (B) Promoter enzyme immunoassay using wild-type and mutated oligonucleotides within the FOXP3 promoter. Bars show mean ± SE of three independent experiments. (C) Chromatin immunoprecipitation assay results show binding of RUNX1 and RUNX3 complexes containing CBFβ to the human FOXP3 promoter in naive CD4+ T cells that were cultured with IL-2 together with anti-CD2/3/28 and TGF-β. There was no change in site occupancy in all immunoprecipitations when IGX1A negative control primers were used. The results are normalized to input and isotype control antibody. Bars show mean ± SE of three independent experiments. Statistical differences were verified by the paired Student's t test. *, P \u003c 0.05\n\nRegulation of FOXP3 promoter activity and FOXP3 protein expression by RUNX1 and RUNX3\nTo investigate the effect of RUNX1 and RUNX3 binding to the RUNX binding sites in the FOXP3 promoter, we transfected human peripheral blood CD4+ T cells with a FOXP3 promoter luciferase reporter vector and RUNX1 or RUNX3 expression vectors. An increase in luciferase activity was observed only when the FOXP3 promoter (−511 to +176) luciferase construct was cotransfected with RUNX1 or RUNX3 expression vectors (Fig. 4 A). The increase in promoter activity was greater upon cotransfection of RUNX3 compared with RUNX1. Luciferase expression was abrogated when the Runx binding sites in the FOXP3 promoter (−511 to +176) luciferase construct were mutated (Fig. 4 A). In these experiments, the overexpression of RUNX1 and RUNX3 eliminated the need of TGF-β for FOXP3 promoter activation and PMA/ionomycin stimulation was sufficient.\nFigure 4. Regulation of FOXP3 promoter activity by RUNX1 and RUNX3. (A) Human primary CD4+ cells were transfected with an empty vector (pGL3 Basic), a vector containing the wild-type or mutated FOXP3 promoter region (FOXP3 −511/+176) fused to the luciferase reporter gene together with a GFP, RUNX1, or RUNX3 expression vector. Bars show the mean luciferase activity ± SE measured as arbitrary light units of three independent experiments. (B) Human primary CD4+ cells were transfected with an empty vector (pGL3 Basic), a vector containing the putative FOXP3 promoter region (FOXP3 −511/+176) fused to the luciferase reporter gene, or with a vector containing the putative FOXP3 promoter region (FOXP3 −511/+176) with single RUNX binding sites mutated (53, 287, or 333) or with the combination of two or three RUNX binding sites mutated (53, 287, or 333) fused to the luciferase reporter gene. Bars show the mean luciferase activity ± SD measured as arbitrary light units of three independent experiments. To examine the role of each of the three RUNX binding sites for the FOXP3 promoter activity, we mutated each individually or in combination. No reduction in luciferase activity was observed when the −53 site was mutated and only a slight reduction when either the −287 or −333 site was mutated (Fig. 4 B). However, mutating the −53 site in combination with one of the other two sites led to a significant decrease in luciferase activity, with the greatest reduction observed when all three binding sites were mutated (Fig. 4 B), suggesting that the identified binding sites have redundant functions and RUNX binding to more than one site is necessary for the full activation of the FOXP3 promoter. Supporting these findings, the overexpression of RUNX1 in human primary CD4+ T cells resulted in significantly elevated levels of FOXP3 protein measured by flow cytometry after 48 h. This was achieved without any requirement for anti-CD3, anti-CD28 stimulation, or the presence of TGF-β. Although there was a trend, the transfection of CD4+ T cells with RUNX3 did not lead to statistically significant increase in FOXP3 (Fig. S5).\n\nRole of CBFβ in the induction of Foxp3\nCBFβ, a common cofactor of all RUNX proteins, stabilizes and increases the binding of the runt domain to target DNA sequences. To target all Runx proteins that might be involved in the induction of Foxp3, we used mice in which loxP-flanked Cbfb alleles were inactivated in T cells through expression of a CD4-cre transgene. Retinoic acid and TGF-β synergize in the induction of Foxp3 in naive T cells (Kang et al., 2007). To investigate whether Runx-mediated induction of Foxp3 is dependent on the expression of CBFβ, naive CD4+ CD8− T cells from CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice were stimulated with anti-CD3/28 mAbs, retinoic acid, and increasing concentrations of TGF-β. After 3 d in culture, the cells were restimulated with PMA and ionomycin and analyzed for intracellular Foxp3 and IFN-γ expression. TGF-β induced Foxp3 in CbfbF/+ CD4-cre cells in a dose dependent manner, and this was significantly reduced in CbfbF/F CD4-cre cells. Retinoic acid enhanced Foxp3 expression even in 20 pg/ml of TGF-β and more than 95% of the CD4+ T cells from CbfbF/+ CD4-cre mice became Foxp3+ in 100 and 500 pg/ml TGF-β doses. The induction of Foxp3 was again significantly lower in CbfbF/F CD4-cre CD4+ T cells even in the presence of retinoic acid, demonstrating that deficiency in Runx binding to DNA affects the TGF-β induction of Foxp3 in T reg cells (Fig. 5 A). There was no difference in the induction of Foxp3 when endogenous IL-4 and IFN-γ were neutralized (Fig. S6).\nFigure 5. Diminished capacity of Cbfb-deficient CD4-cre mice T cells in the generation of Foxp3+ CD4+ T cells. (A) FACS-purified naive CD4+ CD8− T cells from CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice were activated in vitro with anti-CD3/28 mAb, 50 U/ml IL-2, ± 10 nM retinoic acid (RA), and increasing concentrations of TGF-β. After 3 d in culture, the cells were restimulated with PMA + ionomycin, and then analyzed for intracellular Foxp3 and IFN-γ expression. One of five experiments is shown. (B) 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 per transfer). 6 wk later, TCRβ+CD4+ cells from the spleen, mesenteric lymph node (MLN), and lamina propria of the small intestine (LP) were analyzed for Foxp3-GFP expression. Results from one of four CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice with same findings are shown. The data from four sets of mice is shown in C. Statistical analysis was performed with Mann-Whitney U test. *, P \u003c 0.05 between groups. The importance of RUNX transcription factors for the in vivo conversion of naive CD4+ T cells into iT reg cells was examined. Control CbfbF/F or CbfbF/F CD4-cre naive T cells, harboring a Foxp3-IRES-GFP allele were adoptively transferred into Rag2−/− mice. 6 wk later, CD4+ T cells in spleen, mesenteric lymph node, and lamina propria of the small intestine were analyzed for Foxp3-GFP expression (Fig. 5 B). There was a consistently lower percentage of CD4+ T cells that had developed Foxp3 expression in the mesenteric lymph node and lamina propria of mice transferred with CbfbF/F CD4-cre cells compared with control cells (Fig. 5, B and C). These data affirm the significance of RUNX proteins for the in vivo generation of CD4+ Foxp3+ T cells.\n\nCBFβ, RUNX1, and RUNX3 are important for the suppressive activity of CD25+ Foxp3+ T reg cells\nEven though Cbfb-deficient CD4+ T cells had impaired induction of Foxp3 after stimulation with anti-CD3/28 mAbs and TGF-β, sufficient numbers of Foxp3+ cells could be generated to permit analysis of their suppressive activity. Purified naive CD4+ T cells from CbfbF/F CD4-cre and control CbfbF/+ CD4-cre mice (Cd45.2) harboring a Foxp3-ires-GFP allele (Bettelli et al., 2006) were stimulated in vitro with anti-CD3/28 mAbs, IL-2, and TGF-β. After 3 d, Foxp3+-GFP+ cells were sorted by flow cytometry and mixed with CFSE-labeled naive CD45.1+ CD4+ cells at ratios of 1:4, 1:2, and 1:1. The cells were then incubated with inactivated splenocytes and stimulated with anti-CD3 mAb for four more days. CD45.1+ cells were analyzed for CFSE dilution (Fig. 6 A). CbfbF/+ CD4-cre CD4+ T cells activated in the presence of TGF-β showed a clear suppression of T cell proliferation that became even more apparent when an increased ratio of FOXP3+/CD25− cells was used (Fig. 6 B). The suppression was significantly reduced when cells from CbfbF/F CD4-cre mice were used, demonstrating that TGF-β–induced Runx complexes are important for the suppressive activity of Foxp3+ T reg cells (Fig. 6 B). As a control, CbfbF/+ CD4-cre CD4+ T cells activated in absence of TGF-β were mixed with CFSE-labeled naive CD45.1+ CD4+ cells. No suppression could be observed in all control groups without TGF-β at all tested ratios (Fig. 6 C). The decreased suppression capacity of CbfbF/F CD4-cre iT reg cells was unlikely to be caused by decreased survival or proliferation. Foxp3+ and Foxp3− cells generated from both CbfbF/F CD4-cre and control cells all displayed similar proliferation rates and cell death as measured by CFSE dilution and annexin V staining, respectively (Fig. S7, A and B).\nFigure 6. CbfbF/F CD4-cre mouse cells and iT reg cells generated from human naive CD4+ T cells undergoing siRNA-mediated RUNX1 and RUNX3 knock down show a diminished suppressive activity. Experimental setup (A) and results of the mouse suppression assay (B), FACS-purified naive CD4+8− T cells from CbfbF/F CD4-cre (left) and control CbfbF/+ CD4-cre mice (Cd45.2; right) were activated in vitro with anti-CD3/28 mAb, 50 U/ml IL-2, and 2.5 ng/ml TGF-β. After 3 d, Foxp3-GFP+ cells were FACS-sorted and mixed with CFSE-loaded naive CD45.1+ CD4+ cells at the indicated ratios. These were then incubated with inactivated splenocytes and anti-CD3 mAb. After a further 4 d, CD45.1+ cells were analyzed for CFSE dilution. (C) As a control, CbfbF/+ CD4-cre CD4+ T cells activated in absence of TGF-β were mixed with CFSE-loaded naive CD45.1+ CD4+ cells at the indicated ratios. Four days later CD45.1+ cells were analyzed for CFSE dilution. The median division number (of the naive CD45.1+ CD4+ cells in the cultures containing Foxp3-GFP+ cells) is indicated in each of the histograms. One of three experiments is shown. (D) Human naive CD4+ T cells, transfected with RUNX1 and RUNX3 siRNA and cultured under iT reg differentiating conditions were used in an in vitro suppression assay, cultured together with autologous CFSE-labeled CD4+ T cells, and stimulated with anti-CD3 mAb. The CFSE dilution of the CD4+ T cell responder cells was analyzed after 5 d by flow cytometry. The T reg/responder CD4+ T cell ratios used were 1:20, 1:10, and 1:5. One of two experiments is shown. In addition, we tested the requirement of RUNX1 and RUNX3 for the development of the suppressive capacity in human iT reg cells. We isolated human naive CD4+ T cells and transfected them with a combination of RUNX1 and RUNX3 siRNA, or with a scrambled control siRNA. Cells were cultured under T reg conditions, and then mixed with CFSE-labeled autologous CD4+ T cells and stimulated with anti-CD3 mAb. Cells in which RUNX1 and RUNX3 were knocked down showed markedly lower suppressive activity compared with control iT reg cells at a T reg/CD4+ T responder cell ratio of 1:20, but not when the T reg/CD4+ T responder cell ratio was increased to 1:5 (Fig. 6 D). These results demonstrate the important role of RUNX1 and RUNX3 not only for the induction of FOXP3, but also for the suppressive capacity of iT reg cells both in humans and in mice. The data suggest both quantity and quality of T reg cells are hampered. Reduced intrinsic suppressive capacity of iT reg cells was demonstrated in mice, because Foxp3-GFP+ cells were FACS sorted and same numbers of iT reg cells are included in control experiments. In the suppression experiment with human cells, reduced suppressive activity was caused by reduced FOXP3 expression in T cells by siRNA inhibition of RUNX1 and RUNX3.\n"}

bionlp-st-ge-2016-spacy-parsed

UBERON-AE

GO-BP

GO-MF

GO-CC

sentences

events-check-again

bionlp-st-ge-2016-reference-tees

bionlp-st-ge-2016-reference

bionlp-st-ge-2016-uniprot

test2

PMC:2806624 / 5858-26996 JSONTXT

Annnotations TAB JSON ListView MergeView

PMC:2806624 / 5858-26996 JSON TXT