defective cell death signalling along the bcl 2 regulated apoptosis pathway compromises treg cell development and limits their functionality in mice

Journal of Autoimmunity 38 (2012) 59e69 Contents lists available at SciVerse ScienceDirect Journal of Autoimmunity journal homepage: www.elsevier.com/locate/jautimm Defective cell death signalling along the Bcl-2 regulated apoptosis pathway compromises Treg cell development and limits their functionality in mice Denise Tischner a, Irene Gaggl a, Ines Peschel a,1, Manuel Kaufmann a, Selma Tuzlak a, Mathias Drach b, Nikolaus Thuille c, Andreas Villunger a, 2, *, G Jan Wiegers a, 2, * a b c Biocenter, Division of Developmental Immunology, Innsbruck Medical University, A-6020 Innsbruck, Austria Institute of Pathology, Innsbruck Medical University, Innsbruck, Austria Experimental Cell Genetics, Department for Medical Genetics, Molecular and Clinical Pharmacology, Innsbruck Medical University, Innsbruck, Austria a r t i c l e i n f o a b s t r a c t Article history: Received 28 November 2011 Received in revised form 16 December 2011 Accepted 19 December 2011 The Bcl-2 regulated apoptosis pathway is critical for the elimination of autoreactive lymphocytes, thereby precluding autoimmunity T cells escaping this process can be kept in check by regulatory T (Treg) cells expressing the transcription and lineage commitment factor Foxp3 Despite the wellestablished role of Bcl-2 family proteins in shaping the immune system and their frequent deregulation in autoimmune pathologies, it is poorly understood how these proteins affect Treg cell development and function Here we compared the relative expression of a panel of 40 apoptosis-associated genes in Treg vs conventional CD4ỵ T cells Physiological signicance of key-changes was validated using genemodified mice lacking or overexpressing pro- or anti-apoptotic Bcl-2 family members We define a key role for the Bim/Bcl-2 axis in Treg cell development, homeostasis and function but exclude a role for apoptosis induction in responder T cells as relevant suppression mechanism Notably, only lack of the pro-apoptotic BH3-only protein Bim or Bcl-2 overexpression led to accumulation of Treg cells while loss of pro-apoptotic Bad, Bmf, Puma or Noxa had no effect Remarkably, apoptosis resistant Treg cells showed reduced suppressive capacity in a model of T cell-driven colitis, posing a caveat for the use of such longlived cells in possible therapeutic settings Ó 2011 Elsevier Ltd All rights reserved Keywords: Apoptosis Regulatory T cells Bim Bcl-2 Autoimmunity Introduction Regulatory T (Treg) cell dependent suppression and apoptosis of immune effector cells are both essential in establishing and maintaining peripheral tolerance Failure in either process can result in an overshooting immune response and foster the development of autoimmunity Two separate apoptosis signalling pathways contribute to preclude autoimmunity The extrinsic pathway is induced by Abbreviations: TGF-b, transforming growth factor beta; TCR, T cell receptor; Bcl-2, B cell lymphoma 2; Bim/Bcl-2L1, 1Bcl-2 interacting mediator of cell death; MFI, mean fluorescence intensity; Tcon, conventional CD4ỵ T cells; iTreg, induced regulatory T cells; nTreg, natural regulatory T cells; Foxp3, Forkhead box P3; GFP, green fluorescent protein; IL, interleukin; IFN-g, interferon gamma; BH, Bcl-2 homology; GITR, glucocorticoid-induced TNFR-related protein; CTLA-4, cytotoxic T lymphocyte antigen * Corresponding authors Tel.: ỵ43 512 9003 70964; fax: ỵ43 512 9003 73960 E-mail addresses: andreas.villunger@i-med.ac.at (A Villunger), jan.wiegers@imed.ac.at (G Jan Wiegers) Current address: Division of Medical Biochemistry, Biocenter, Innsbruck Medical University, Innsbruck, Austria Equal contribution 0896-8411/$ e see front matter Ó 2011 Elsevier Ltd All rights reserved doi:10.1016/j.jaut.2011.12.008 ligation of membrane-bound death receptors (DR; e.g CD95 or TRAIL-R) with their cognate ligands (CD95L or TRAIL), whereas the intrinsic or mitochondrial pathway is triggered in response to diverse forms of cell stress including cytokine-deprivation or highaffinity ligation of antigen receptors and mediated by proteins of the Bcl-2 family [1] The latter are divided according to function in anti- and pro-apoptotic proteins Anti-apoptotic proteins (e.g Bcl-2, Bcl-x or Mcl-1) contain up to four Bcl-2-homology domains (BH14), pro-apoptotic Bax/Bak-like proteins three (BH1-3) and ‘BH3only’ proteins (e.g Bim, Puma) only one, i.e the BH3 domain Defective signalling along the extrinsic apoptosis pathway can underlie the pathogenesis of autoimmune disease in mice and humans as exemplified by the loss of CD95-mediated apoptosis in autoimmune prone lpr mice and patients suffering from autoimmune lymphoproliferative syndrome (ALPS) [2] Also, deviations of the intrinsic pathway contribute to the establishment of autoimmunity Loss of Bim or Bcl-2 overexpression in mice impairs negative selection of thymocytes and developing B cells expressing selfreactive antigen receptors and Bim plus Puma co-regulate lymphocyte homeostasis in the periphery [3] Forced overexpression of Bcl-2 or Mcl-1 causes lymphadenopathy, can facilitate 60 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 cancer development [4,5] and the former also facilitates autoimmunity in mice [6,7] Deletion of activated cells after antigenic challenge is impaired in Bim-deficient or Bcl-2 overexpressing animals thereby facilitating the development of SLE-like pathology [7,8] Consistently, high-level expression of Bcl-2 or its pro-survival homologues is frequently associated with different types of AID in humans [9e11] Treg cells are characterized by the expression of distinct cell surface molecules including CD4, the IL-2Ra chain (CD25), GITR and CTLA-4 but the transcription factor Foxp3 appears to be the only reliable marker [12] Treg cells arise naturally in the thymus (nTreg) or can be induced (iTreg) in the periphery from CD4ỵFoxp3 naùve T cells in response to TGF-b plus IL-2 or retinoic acid [13] Their immune suppressive capacity involves the secretion of antiinflammatory cytokines (e.g TGF-b, IL-10), cell-to-cell contact dependent mechanisms (e.g CTLA-4) or active transfer of small immune-modulatory metabolites such as cAMP [14,15] Previous reports also suggested that induction of apoptosis in activated T cells, e.g due to expression of DR ligands such as TRAIL [16] or CD95L [17] on Treg cells, or Treg-dependent IL-2 deprivation of activated T cells [18], triggering Bcl-2 regulated cell death, may contribute Loss of Foxp3 triggers the scurfy phenotype in mice and immune dysregulation, polyendocrinopathy, enteropathy, and X-linked inheritance (IPEX) in humans [19,20] Importantly, the timed deletion of Treg cells in adult mice results in scurfy-like pathology [21] Moreover, Treg number and function are frequently reduced in patients suffering from autoimmunity [22,23] These findings demonstrate the importance and simultaneously highlighting the therapeutic potential of Treg cells for the treatment of autoimmunity or other pathologies such as graft vs host disease Despite the well-established role of Bcl-2 family proteins in lymphocyte development and homeostasis, only fragmented information is available concerning their impact on Treg cell biology To gain insight, we investigated the relative expression levels of Bcl-2 family proteins in Treg cells derived from thymus or spleen in relation to that found in developing thymocytes or conventional CD4ỵ T cells In addition, we explored the impact of loss- or gain-of-function of key Bcl-2 family proteins on Treg cell development, their cell death responsiveness as well as their suppressor function in vitro and in vivo fluorochrome-labelled antibodies recognizing mouse CD4 and 7AAD to exclude dead cells Cell sorting was performed using a FACSVantage cell sorter (Becton Dickinson) Purity of isolated cell populations was routinely !98% 2.3 Flow cytometry The following fluorochrome-labelled antibodies or reagents were used for extra- and intracellular staining: rat anti-mouse CD4 mAb (L3T4), rabbit anti-mouse GITR (YGITR 765, Biolegend), rat antimouse Foxp3 (FJK-16s) from eBioscience; 7AAD from Sigmae Aldrich; rat anti-mouse CD25 mAb (3C7), hamster anti-mouse CTLA-4 mAb (UC10-4B9) and AnnexinV-Alexa647, rat anti-mouse IFN-g (XMG1.2), rat anti-mouse IL-17A mAb (TC11-18H10.1) from Biolegend For intracellular staining of cytokines cells were stimulated with 50 ng/ml PMA (Fluka Biochemika) and mg/ml Ionomycin (Sigma) for h During the last h of cell culture Monensin (Biolegend) was added Then cells were fixed with fixation buffer and permeabilized with permeabilization buffer (Biolegend) according to the manufacturer’s instructions For intracellular Foxp3 staining, buffers were purchased from eBioscience Flow cytometry measurements were performed using a FACSCalibur (BD Biosciences) and analyzed by CellquestÔ (BD Biosciences) or WinMDI 2.4 Apoptosis assays To asses apoptosis susceptibility cells were cultured in media (supplemented with 10% FCS, 100 U/ml Pen/Strep, mM glutamine, mM pyruvate, 1 non-essential amino acids, 50 mM betamercaptoethanol) in the presence or absence of 100 U/ml IL-2 (Peprotech), 20 ng/ml IL-7 (Peprotech), 10À8 M Dexamethasone (SigmaeAldrich), 10 mg/ml Etoposide (SigmaeAldrich), mM SAHA (a gift from R Johnstone, Peter MacCallum Cancer Center, Melbourne, Australia), 100 nM Staurosporine (SigmaeAldrich) To induce apoptosis by Fas ligation, cells were cultured in the presence of human 100 ng/ml FasL (Alexis Biochemicals) and crosslinked with mg/ml anti-FLAG (M2, SigmaeAldrich) The percentage of living cells was assessed by AnnexinV/7AAD staining Increased and relative survival was calculated by normalisation to medium cultured cells Materials and methods 2.5 T cell suppression assay 2.1 Mice Freshly puried CD4ỵFoxp3-GFPỵ nTreg cells were tested functionally in a co-culture suppression assay Triplicates of different nTreg cell ratios were co-cultured with 104/ml splenic CD4ỵFoxp3-GFP responder T cells,  105/ml APCs (irradiated total splenocytes, 30 Gy) and 0.5 mg/ml anti-CD3 (2C11) in 96-well U-bottom plates for 72 h mCi [3H]-thymidine/well was added for the last 16 h Cells were transferred onto glass fiber filters using a Combi-cell-harvester (Molecular Devices) and proliferation measured by scintillation counting in a beta-Counter (Beckman Coulter) C57BL/6 foxp3gfp reporter mice [24] were purchased from Jackson Laboratories, backcrossed more generations to C57BL/6 (total times) and crossed with congenic BimÀ/À [8], vav-Bcl-2 [25] mice to obtain foxp3gfpBimÀ/À and foxp3gfpvav-Bcl-2 mice RAG1À/À mice were a kind gift from A Moschen, Department of Internal Medicine II The generation of mice deficient for Bad [26], Bmf [27], Puma [28] or Noxa [28] has been described All animals used in this study were on a C57BL/6 background and 6e12 weeks of age All animal experiments were performed in accordance with the Austrian “Tierversuchsgesetz” (BGBl Nr 501/1988 i.d.F 162/2005) and have been granted by the Bundesministerium für Bildung, Wissenschaft und Kultur (BMWF-66.011/0167-II/3b/2011) 2.2 Cell sorting To obtain CD4ỵFoxp3-GFP conventional T cells and CD4ỵFoxp3GFPỵ natural Treg cells single cell suspensions were prepared from splenocytes and thymocytes isolated from foxp3gfp wild type, foxp3gfpBimÀ/À, foxp3gfpvav-Bcl-2 mice and stained with 2.6 RNA isolation and quantitative RT-PCR RNA isolation (Zymo research) and cDNA synthesis (Biorad) were performed according to the manufacturer’s instructions Real time PCR was done using the following primers (50 e30 ): A1 CCTGGCTGAGCACTACCTTC (sense) and TCCACGTGAAAGTCATCCAA (antisense); b-actin ACTGGGACGACATGGAGAAG (sense) and GGGGTGTTGAAGGTCTCAAA (antisense); Bcl-2, CTGGCATCTTCTCC TTCCAG (sense) and GACGGTAGCGACGAGAGAAG (antisense); BclxL TTCGGGATGGAGTAAACTGG (sense) and TGGATCCAAGGCTCT D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 AGGTG (antisense); Bim GAGATACGGATTGCACAGGA (sense) and TCAGCCTCGCGGTAATCATT (antisense); Mcl-1 TAACAAACTGGGGCA GGATT (sense) and GTCCCGTTTCGTCCTTACAA (antisense); Noxa CCCACTCCTGGGAAAGTACA (sense) and AATCCCTTCAGCCCTTGATT (antisense); Puma CAAGAAGAGCAGCATCGACA (sense) and TAGTTGGGCTCCATTTCTGG (antisense) Quantitative RT-PCR was performed using a MastercyclerÒ Gradient (Eppendorf) and the DyNAmoÔ Flash SYBR mastermix (Finnzymes) according to the manufacturer’s instruction The results were normalized to b-actin expression and evaluated using the ÀDDCT relative quantification method 2.7 Multiplex ligation-dependent probe amplification (RT-MLPA) RNA amounts were analyzed with the Apoptosis Mouse RTMLPA kit RM002 (MRC-Holland, http://www.mlpa.com) according to the manufacturer’s instructions [29] Samples were run through a Genescan and analyzed with Gene-Mapper (Applied Biosystems GmbH; http://www.appliedbiosystems.com) and subsequently using the StatviewÓ software program (Abacus) 2.8 T cell transfer model of colitis CD4ỵFoxp3-GFP conventional T cells were isolated from congenic C57BL/6 foxp3gfp mice and injected i.p into 6e10-week-old C57BL/6 RAG1À/À immunodeficient recipients (4  105 cells/ mouse) Â105 wild type or vav-Bcl-2 CD4ỵFoxp3-GFPỵ Treg cells were co-injected i.p where indicated Mice were monitored every 3e4 days for wasting disease Mice were sacrificed either losing >25% of its initial body weight or weeks after cell transfer 2.9 Histological assessment of intestinal inflammation Samples of proximal colon, mid-colon, and distal colon were fixed in buffered 4% formalin solution mm paraffin-embedded sections were cut and stained with hematoxylin and eosin Tissues were graded semi-quantitatively from grade to in a blinded fashion Grade 0: no changes observed, grade 1: discrete increased inflammatory cells in the lamina propria with granulocytes in the lamina epithelialis, grade 2: as grade with scattered erosions of the mucosa, grade 3: increased inflammatory cells in the lamina propria and scattered crypt abscesses, grade 4: all signs of grade plus more than crypt abscesses per colon circumference in the scanning magnification (40Â) 2.10 Statistics Estimation of statistical differences between groups was carried out using the unpaired Student’s t-test or Mann Whitney U test, where appropriate P-values of 0.05 were considered to indicate statistically significant differences Results 3.1 Distinct expression patterns of anti- and pro-apoptotic Bcl-2 family members are found in nTreg vs conventional T cells First we investigated if Treg and conventional CD4ỵ T cells (Tcon) differ in their repertoire of 40 apoptosis-associated genes, including all pro- and anti-apoptotic Bcl-2 family members Therefore, Treg cells, CD4ỵ8 thymocytes or CD4ỵ Tcon cells were isolated from the thymus or spleen of foxp3gfp reporter mice based on cell surface marker and GFP expression by cell sorting RNA was isolated and analyzed by RT-MLPA (Suppl Table 1) and quantitative (q)RT-PCR (Fig 1A) In the thymus, Treg cells expressed higher 61 mRNA levels encoding anti-apoptotic Mcl-1, Bcl-xL or A1/B-1, when compared to CD4ỵ8 thymocytes In contrast, expression of pro-apoptotic Bim was found reduced in Treg cells while mRNA for Puma was more and that for Noxa seemed most abundant in this relative comparison (Fig 1A) In contrast, in the spleen only mRNAs for Bcl-xL, A1/Bfl-1 and Noxa were found to be overrepresented in Treg cells, whereas no difference was observed for transcripts encoding Bcl-2, Mcl-1, Bim or Puma, suggesting that Bcl-xL levels may define increased cell death thresholds in nTreg cells in the periphery All other cell death related genes amplified in the RT-MLPA analysis were either not expressed or found comparable in their levels with one notable exception, BIRC1A/NAIP, a member of the NOD-like receptor family, initially implicated as an apoptosis inhibitor in neurons, that was significantly more abundant in Treg cells derived from thymus or spleen (Suppl Table 1) Next we investigated if loss of BH3-only protein function or Bcl2 overexpression had an impact on nTreg cell homeostasis in vivo Therefore, we quantified the percentage and number of CD4ỵFoxp3ỵTreg cells in the thymus and spleens of mice lacking the BH3-only proteins Bim, Bmf, Bad, Puma or Noxa and compared it to the consequences of transgenic overexpression of Bcl-2 This analysis revealed a relative increase in the percentage of nTreg cells in Bim-deficient or vav-Bcl-2 transgenic mice (Fig 1B) No differences in Treg cell number were observed in all other knockout animals analyzed (Fig 1C), including those lacking Puma or Noxa, despite the higher mRNA levels observed in wild type Treg cells (Fig 1A) 3.2 nTreg cells differ in their cell death responsiveness from conventional T cells Based on our findings related to gene expression and the observed accumulation of Treg cells in Bim-deficient or Bcl-2 overexpressing mice, we assessed if Treg cells differ from CD4ỵ8 thymocytes or CD4ỵ Tcon cells from spleen in their susceptibility to apoptosis Therefore, CD4ỵFoxp3-GFPỵ Treg cells, CD4ỵ8Foxp3GFP thymocytes or CD4ỵFoxp3-GFP Tcon cells were isolated from the thymus or spleen of mice from the different genotypes Cells were exposed to a broad range of apoptotic stimuli triggering Bcl-2 dependent apoptosis, including cytokine deprivation, treatment with the glucocorticoid dexamethasone, the DNA-damaging drug etoposide, the histone-deacetylase (HDAC)-inhibitor SAHA, the broad-spectrum kinase-inhibitor staurosporine and, for reference, CD95 ligation Compared to CD4ỵ8 thymocytes (Suppl Fig 1) or splenic Tcon cells (Fig 2), Treg cells died more rapidly in the absence of cytokines and were efficiently rescued by addition of IL2, in line with their strict dependence on this cytokine for survival [30], but also by the addition of IL-7 Consistently, loss of Bim as well as Bcl-2 overexpression protected them from spontaneous death in culture (Fig 2AeC) Next to cytokine deprivation, Treg cells were found more susceptible to apoptosis induced by the HDAC inhibitor SAHA, and Fas ligation (Fig 2D, E) Loss of Bim or Bcl-2 overexpression protected Treg cells from SAHA but not CD95-induced apoptosis In contrast, Treg cells were more resistant to apoptosis induced by the glucocorticoid dexamethasone, etoposide or staurosporine when compared with Tcon cells (Fig 2FeH) This difference might be explained with increased level of Bcl-xL in these cells (Fig 1), whereas in the thymus, Treg cells displayed increased survival only in response to staurosporine treatment As in Tcon cells, absence of Bim or Bcl-2 overexpression conferred partial or complete protection to dexamethasone or staurosporine while etoposide killing was only blocked by Bcl-2 overexpression (Fig 2; Suppl Fig 1) 62 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 relative mRNA expression spleen 10 ** *** *** ** *** *** relative mRNA expression thymus A 10 ** * * l-2 l-1 xL Bc Mc Bcl- B A1 Bim oxa uma N P l-2 l-1 xL Bc Mc Bcl- A1 Bim oxa uma N P thymus spleen ** 15 10 ** l wi C Treg among CD4+8thymocytes [%] Treg among CD4+ T cells [%] *** e yp dt -/Bim 15 10 va clv-B t ld wi thymus e -/-/-/-/typ Bmf uma oxa Bad ld N i P w /mBi clv-B a v 15 15 10 e yp spleen Treg among CD4+ T cells [%] Treg among CD4+8thymocytes [%] ** 10 e -/-/-/-/typ Bmf uma oxa Bad ld N i P w Fig Differential expression of proteins of the Bcl-2 family by Treg cells in the thymus and spleen CD4ỵFoxp3-GFPỵ Treg, CD4ỵ8Foxp3-GFP thymocytes, or CD4ỵFoxp3-GFP Tcon cells were isolated from the (A) thymus or spleen, respectively Relative mRNA expression of Bcl-2 family members was analyzed by quantitative RT-PCR Expression in CD4ỵFoxp3-GFP Tcon cells was set to Bars represent means ặ SEM of n ẳ independent samples (B) Relative abundance of CD4ỵFoxp3-GFPỵ Treg cells among CD4ỵ T cells in the thymi and spleens of foxp3gfpwt (n ¼ 8), foxp3gfpBimÀ/À (n ¼ 3) and foxp3gfpvav-Bcl-2 (n ¼ 5) mice (C) Abundance of CD4ỵFoxp3-GFPỵ Treg cells among CD4ỵ T cells in the thymi and spleens of wt (n ! 8), BmfÀ/À (n ! 4), PumaÀ/À (n ! 5), NoxaÀ/À (n ¼ 2) and BadÀ/À (n ¼ 2) mice Bars represent means Ỉ SEM; statistics: Student’s t-test and Mann Whitney U test for Noxa in the spleen *p 0.05, **p 0.01, ***p 0.001 Collectively, our findings show that although cell death signalling pathways and responses are generally conserved between Tcon and Treg cells, differences in relative sensitivity to certain forms of stress exist, that correlate with the different expression patterns of certain Bcl-2 family proteins described above (Fig 1) 3.3 Reduced expression of Treg cell markers due to loss of Bim or Bcl-2 overexpression Because apoptosis differently affected Treg and Tcon cells we wanted to know if and how deregulation of cell death impacts on Treg cell phenotype and function Therefore, we analyzed expression D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 A Treg B Tcon wt *** *** 40 *** 20 ** ** 80 *** 60 40 20 18 48 120 80 100 60 40 20 Dexamethasone 80 ** ** +++ +++ +++ +++ 40 20 e -2 -/typ Bim -Bcl ld v wi va 20 60 ++ ** 40 20 G Etoposide *** 100 80 60 40 20 ld wi H e -2 -/typ Bim -Bcl v a v Staurosporine 120 80 * 40 20 FasL +++ +++ 100 60 e -2 -/typ Bim -Bcl v a v 120 e -2 -/typ Bim -Bcl ld v i a w v relative survival [%] +++ +++ +++ 100 60 80 +++ +++ +++ 120 120 40 E e -2 -/typ Bim -Bcl ld v i a w v relative survival [%] SAHA 100 relative survival [%] increased survival [%] D IL-7 60 ld wi time [h] time [h] C 48 relative survival [%] 18 Treg Tcon 80 0 F ** increased survival [%] *** *** 60 IL-2 100 100 relative survival [%] 80 -/Bim vav-Bcl-2 survival [%] survival [%] 100 63 e -2 -/typ Bim -Bcl ld v i w va ++ +++ ++ +++ 100 80 * 60 40 20 t ld wi e yp -/- cl-2 Bim v-B va Fig Splenic Treg and Tcon cells display different susceptibility to mitochondrial apoptosis Wild type, Bim/ and vav-Bcl-2 CD4ỵFoxp3-GFPỵ Treg and CD4ỵFoxp3-GFP Tcon cells were purified from the spleen and survival analyzed by AnnexinV/7AAD staining Only AnnexinV/7AAD negative cells were considered alive Cells were cultured either (A) in medium for 9, 18 and 48 h or in the presence of (B) 100 U/ml IL-2, (C) 20 ng/ml IL-7, (D) mM SAHA, (E) 100 ng/ml FasL, (F) 10À8 M Dexamethasone, (G) 10 mg/ml Etoposide and (H) 100 nM Staurosporine for 18 h For calculation of increased and relative survival cell viability was normalized to medium values Symbols and bars represent means Ỉ SEM of n ¼ 3e10 data points acquired in !3 independent experiments; statistics: Student’s t-test *p 0.05, **p 0.01, ***p 0.001 Treg cells were compared to Tcon cells; ỵỵp 0.01, ỵỵỵ p 0.001 wild type Treg and Tcon cells, respectively were compared to their BimÀ/À and vav-Bcl-2 counterparts of Treg cell markers in foxp3gfpwild type, foxp3gfpBimÀ/À and foxp3gfpvav-Bcl-2 mice, the latter two showing increased Treg cell numbers While in wild type mice nearly all Foxp3-GFPỵ cells were CD4ỵ (Fig 3A, B), we observed an increased number of CD8ỵFoxp3GFPỵ cells in the spleens of foxp3gfpBimÀ/À and foxp3gfpvav-Bcl-2 mice Surprisingly, up to 40% of Bim/ and vav-Bcl-2 Foxp3-GFPỵ cells in the thymus lacked CD4 or CD8 expression and even migrated to the periphery (Fig 3A, B) Furthermore, expression of typical Treg cell markers, such as Foxp3, CD25 and GITR were diminished in BimÀ/À mice, an effect even more pronounced, in vav-Bcl-2 Treg cells (Fig 3CeF and Suppl Fig 2) CTLA-4 expression, required for Treg cell effector function, was significantly reduced in thymic BimÀ/À and vav-Bcl-2 Treg cells but this phenomenon was no longer seen in the spleen Overall, Treg cell marker reduction was more pronounced in the thymus than the spleen Together this indicates, that defective cell death signalling impacts on the maturation program of Treg cells that may also impact on their effector function 3.4 Treg cells derived from foxp3gfpBimÀ/À or foxp3gfpvav-Bcl-2 mice show impaired suppressive capacity Expression levels of Foxp3 in Treg cells highly correlate with their suppressive capacity and forced expression of Foxp3 in former Foxp3CD4ỵ T cells confers suppressive function [31] Because Foxp3 amounts were reduced in foxp3gfpBimÀ/À and foxp3gfpvav-Bcl-2 Treg cells we assessed if their suppressive potential might be impaired in an in vitro suppression assay system Therefore, CD4ỵFoxp3-GFP Tcon cells were stimulated either alone or in the presence of genotype matched Treg cells Treg cells from all genotypes were able to suppress their matching Tcon cells but while suppression of wild type or BimÀ/À Tcon cells seemed similar, that of vav-Bcl-2 Tcon cells was clearly less efficient (Fig 4A) This marked difference may be related to changes in function or responsiveness of nTreg and/or Tcon cells overexpressing Bcl-2 that may include increased survival (Fig 2) and/or inferior proliferative capacity (not shown) It has been proposed that responder T cells can differ in their susceptibility to Treg cell mediated suppression when cell death is impaired [18] Thus, we cultured wild type, Bim-deficient or vavBcl-2 transgenic Tcon cells together with wild type Treg cells and assessed their proliferation capacity Proliferation of BimÀ/À and vav-Bcl-2 Tcon cells was efficiently suppressed in the presence of wild type Treg cells (Fig 4B) Surprisingly, suppression of BimÀ/À Tcon cells by wild type Treg cells was even more efficient than that of wild type or vav-Bcl-2 Tcon cells when limiting dilution experiments were performed While the reasons for the increased susceptibility of Bim-decient CD4ỵ T cells over those expressing Fig Treg cell marker expression is reduced in BimÀ/À and vav-Bcl-2 Treg cells (A) Representative dot-plot analysis of CD4 and CD8 cell surface marker expression by Foxp3-GFPỵ Treg cells in the thymus (left column) or spleen (right column) of wild type, BimÀ/À or vav-Bcl-2 mice (B) Quantification of CD4 and CD8 distribution among of Foxp3-GFPỵ Treg cells in the thymus (upper panel) or spleen (lower panel) Bars represent means Ỉ SEM of wt n ¼ 11; BimÀ/À n ¼ 6; vav-Bcl-2 n ¼ animals (C) Representative dot-plot analysis of CD25 expression on the cell surface of wild type, BimÀ/À and vav-Bcl-2 Treg cells in the thymus (left column) or spleen (right column) (D) Quantification of CD25 expression as in (B) Mean fluorescence intensity (MFI) of Foxp3-GFP, GITR and CTLA-4 by CD4ỵFoxp3-GFPỵ Treg cells in the (E) thymus or (F) spleen (wt n ! 9; BimÀ/À n ! 3; vav-Bcl-2 n ! 3); bars represent means Ỉ SEM of !3 independent experiments; statistics: Student’s t-test *p 0.05, **p 0.01, ***p 0.001 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 transgenic Bcl-2 remain to be identified, our observations demonstrate that induction of apoptosis in Tcon cells does not play a significant role in suppression, as most recently also noted by Vignali et al [32] Keeping in mind that Bim-deficient T cells were reported to be more anergic and to respond more reluctantly to antigenic challenge [33], we reasoned that differences in Treg cell function may become distinguishable when apoptosis-defective Treg cells needed to suppress functionally fully competent responder T cells In line with this hypothesis, when wild type Tcon cells were incubated with BimÀ/À or vav-Bcl-2 Treg cells a clear-cut difference was observed, as both genotypes were less effective in suppressing wild type Tcon cells, when compared to controls (Fig 4C) The suppressive capacity of vav-Bcl-2 Treg cells was weakest and this phenotype correlated with lower levels of Foxp3 in these cells, a phenomenon also noted in BimÀ/À Treg cells, albeit less pronounced (Fig 3) To assess if Foxp3 expression directly impacts on Treg cell function we divided vav-Bcl-2 Treg cells in Foxp3GFPhigh and Foxp3-GFPlow groups and compared their suppressive potential with that of unseparated (Foxp3-GFPall) and wild type Treg cells that all showed high levels of GFP expression (Foxp3GFPhigh) (Fig 4C) Indeed, Foxp3-GFPlow vav-Bcl-2 Tregs were weaker in the suppression of Tcon cells compared to all other Treg subpopulations Surprisingly, Foxp3-GFPhigh vav-Bcl-2 Treg cells were still less potent when compared to wild type Foxp3-GFPhigh Treg cells (Fig 4D), indicating that the levels of Foxp3 in these cells not define suppressive capacity alone Impaired function becomes best apparent when apoptosis-defective Treg cells are challenged with fully competent CD4ỵ responder T cells Hence deregulation of Bcl-2 family protein levels in Treg cells may contribute to inflammatory phenotypes frequently underlying autoimmunity 65 3.5 Bcl-2 transgenic Treg cells fail to effectively suppress inflammatory bowel disease To evaluate if vav-Bcl-2 derived Treg cells also display an impaired suppressive function in vivo we tested their capacity to prevent T cell induced colitis Therefore,  105 CD4ỵFoxp3-GFP wild type conventional T cells were transferred alone (control group) or together with 105 CD4ỵFoxp3-GFPỵ wild type or vavBcl-2 Treg cells into RAG1À/À mice Animals receiving only Tcon cells without Treg cells rapidly developed colitis, indicated by strong weight loss and had to be sacrificed early on, displaying a colitis score of 3, according to histopathological assessment (Fig 5A; not shown) Transfer of wild type Treg cells successfully prevented development of colitis during the whole observation period whereas mice receiving vav-Bcl-2 Treg cells succumbed to disease, initially even as quickly as mice that only received conventional CD4ỵ T cells (Fig 5A) However, after two weeks their weight loss decelerated, suggesting at least a partial suppression of disease by vav-Bcl-2 Treg cells but animals never fully recovered during the whole observation period (Fig 5AeC) Animals were sacrificed after weeks Consistent with a pro-inflammatory phenotype, total splenic cellularity was significantly increased in animals that received vav-Bcl-2 Treg cells compared to wild type controls due to an increase in Mac-1ỵ cells (not shown), while the number of conventional CD4ỵ T and Treg cells remained comparable (Suppl Fig 3A) Notably, significantly more IFN-g producing CD4ỵ Tcon cells were found in the mesenteric lymph nodes of mice treated with Treg cells overexpressing Bcl-2 than mice that received wild type cells while no difference was observed in IL-17 production (Fig 5E, F) Mesenteric lymph node cellularity (Fig 5D) and IFN-g or IL-17 secretion in spleen were indistinguishable between the groups (Suppl Fig 3B) Together, these findings strongly support Fig BimÀ/À and vav-Bcl-2 Treg cells display reduced suppressive capacity in vitro Different ratios of splenic CD4ỵFoxp3-GFPỵ Treg and CD4ỵFoxp3-GFP Tcon cells were stimulated with anti-CD3 mAb (2C11) for days T cell proliferation was assessed by [H3]-thymidine incorporation added during the last 16 h of cell culture Proliferation was normalized to Tcon cell proliferation in the absence of Treg cells (A) Proliferation of wild type (n ¼ 19), BimÀ/À (n ¼ 6) or vav-Bcl-2 (n ¼ 9) Tcon cells in the absence or presence of Treg cells with the same genotype (B) Proliferation of wild type (n ¼ 19), BimÀ/À (n ¼ 7) or vav-Bcl-2 (n ¼ 9) Tcon cells cultured in the absence or presence of wt Treg cells (C) Wild type Tcon cells were used as responder T cells and cultured with or without wild type (n ¼ 19), BimÀ/À (n ! 6) or vav-Bcl-2 (n ¼ 9) Treg cells (D) wt Tcon cells were cultured alone or with wt Foxp3-GFPhigh (n ¼ 4), unseparated vav-Bcl-2 (Foxp3-GFPall) (n ¼ 4) Tregs or vav-Bcl-2 Tregs divided into Foxp3-GFPhigh (n ¼ 3) and Foxp3-GFPlow Tregs cells (n ¼ 4) Data points represent means Æ SEM; statistics: Student’s t-test *p 0.05, **p 0.01, ***p 0.001 66 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 Fig vav-Bcl-2 Treg cells fail to effectively prevent induction of colitis Colitis was induced in RAG1À/À mice by adoptive transfer of  105 wild type CD4ỵFoxp3-GFP Tcon cells without (control group; n ẳ 3) or together with  105 wild type (n ¼ 4) or vav-Bcl-2 CD4ỵFoxp3-GFPỵ Treg cells (n ẳ 4) Animals were sacrificed when weight loss reached >25% (control group) or on day 46 after cell transfer (mice receiving either wild type or vav-Bcl-2 Treg cells) (A) Depicted is the relative weight change normalized to the starting weight before cell transfer (B) Representative H&E stained mid-colon sections and (C) colitis score of mice treated with wild type Tcon cells, either together with wild type or vav-Bcl-2 CD4ỵFoxp3-GFPỵ Treg cells (D) Absolute cell number (left panel), Tcon (middle panel) and Treg cell number (right panel) in the mesenteric lymph nodes of mice receiving wild type Tcon cells either together with wild type (wt) or vav-Bcl-2 (Bcl-2) Treg cells Expression of (E) IFN-g and (F) IL-17A by CD4ỵFoxp3-GFP Tcon cells in the mesenteric lymph node after h stimulation in vitro in the presence of PMA/Ionomycin has been analyzed by intracellular staining and flow cytometry Data points represent means Ỉ SEM; one out of two independent experiments is shown; statistics: Student’s t-test *p 0.05, **p 0.01 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 the pathophysiological relevance of deregulated mitochondrial cell death signalling in Treg cell function, warranting a more detailed analysis of Bcl-2 family protein expression in Treg cells from patients with autoimmune disease differences in their susceptibility to apoptosis induced by different stimuli (Figs and 2) Proteins of the Bcl-2 family are critically involved in T cell selection and maturation processes in the thymus For example Mcl-1 and A1 are induced upon cytokine and (pre)TCR signalling in immature double negative thymocytes [34,35] and Bcl-xL becomes the major survival protein in double-positive thymocytes where Bcl-2 levels become repressed [36] and Bim executes negative selection [37] Contrary to conventional CD4ỵ T cells, Treg cells require a stronger TCR signal for development in the thymus and are more resistant to negative selection compared to conventional CD4ỵ T cells [38,39] This idea is also supported by our results Levels of Bcl-xL were found higher and those for Bim were reduced, circumventing negative selection Higher Mcl-1 and A1 levels may result from strong TCR signalling received during negative selection and may be required to antagonize the simultaneous increase in Noxa levels observed in these cells (Fig 1) In the periphery Treg cells also differ from Tcon cells in expression levels of proteins of the Bcl-2 family Most notably, BclxL that is usually downregulated in mature T cells and replaced by Bcl-2 [36] was found increased, indicating either stimulation by self-peptide ligands in the periphery and/or cytokine stimulation by IL-2 via constitutive expression of CD25 on Treg cells These differences may also account for the observed increased survival of splenic Treg cells in response to dexamethasone, etoposide or staurosporine At least for the observed glucocorticoid resistance a former study correlated augmented Bcl-2 levels and IL-2 with survival advantages in Treg cells in response to dexamethasone treatment [40] In general, loss of pro-apoptotic Bim or overexpression of anti-apoptotic Bcl-2 increased survival of both Treg and Tcon cells alike, indicating conserved function of the Bcl-2 proteins family in both cell types Next to the increased survival capacity in response to the aforementioned stimuli Tregs were more susceptible to cytokine deprivation, HDAC-inhibition and Fas ligation As Treg cells cannot produce IL-2 [30] or IL-7 [41] they are highly susceptible to cytokine deprivation Thus, either loss of Bim or Bcl-2 overexpression, as well as addition of common g-chain cytokines IL-2 or IL-7, rescued Treg cells from apoptosis (Fig 2) Former studies reported that 3.6 Inflammation can restore the suppressive capacity of Bcl-2 overexpressing Treg cells To assess which impact the ongoing inflammatory response had in relation to the delayed recovery of mice that received Bcl-2 transgenic Treg cells, we also compared expression of Foxp3-GFP, CD25, GITR and CTLA-4 in wild type and vav-Bcl-2 Treg cells before and after transfer Treg cells in the spleen of vav-Bcl-2 mice displayed significantly reduced levels of Foxp3-GFP, CD25 and GITR (Fig 3), but surprisingly the opposite phenotype was observed weeks after disease induction (Fig 6A, B) While GITR was now comparable between wild type and vav-Bcl-2 Treg cells (Suppl Fig 4D, E), Foxp3 as well as CD25 expression were even significantly higher in vav-Bcl-2 Treg cells (Fig 6A, B and Suppl Fig 4A, B) In addition, when these cells were tested in an in vitro suppression assay, wild type and vav-Bcl-2 Treg both efficiently suppressed wt Tcon cells (Suppl Fig 4F) This indicates that reduced Treg cell lineage marker expression and suppression capacity by vav-Bcl-2 Treg cells not result from a cell intrinsic developmental defect as these deficiencies can be overcome under pro-inflammatory conditions Notably here, as in vav-Bcl-2 mice, we also observed the emergence of CD4Foxp3-GFPỵ vav-Bcl-2 Treg cells after transfer into RAG1À/À mice (Fig 6C and Suppl Fig 4C) This suggests that defective cell death signalling allows their undesired survival after exhaustion that associates with phenotypic changes Discussion In this study, we assessed how Bcl-2 regulated apoptosis impacts on Treg cell apoptosis susceptibility, maturation, phenotype and functionality Our RT-MLPA expression analysis demonstrates that Treg cells differ from Tcon cells in their repertoire of pro- and anti-apoptotic proteins expressed that contribute to B Foxp3-GFP * 25 20 15 10 wt Bcl-2 MFI Foxp3-GFP 30 CD4 cell number CD25 MFI CD25 C cell number Treg wt Treg vav-Bcl-2 160 140 120 100 80 60 40 20 ** wt Bcl-2 99.8% 96.7% 0.2% 3.3% Foxp3-GFP CD4- Treg cells [%] A 67 3.0 ** 2.5 2.0 1.5 1.0 0.5 0.0 wt Bcl-2 Fig Bcl-2 overexpressing Treg cells express higher levels of Foxp3 and CD25 than wild type Treg cells under inflammatory conditions On day 46 after cell transfer mice were sacrificed and expression of (A) CD25, (B) Foxp3-GFP and (C) CD4 by wild type and vav-Bcl-2 Treg cells in the mesenteric lymph nodes was assessed by flow cytometry For the quantification of CD25 and Foxp3-GFP expression, cells were gated on CD4ỵFoxp3-GFPỵ and for CD4 expression on total Foxp3-GFPỵ Treg cells In the upper panel representative histograms or dot plots from one out of four representative stainings are depicted The lower panels display quantification of CD25 MFI, Foxp3-GFP MFI and CD4À Treg cells Bars represent means Ỉ SEM of n ¼ animals per group; statistics: Student’s t-test *p 0.05 **p 0.01 68 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 HDAC-inhibitors induce Treg cell maturation in vivo Increased acetylation of Foxp3 increased its binding to the IL-2 promoter thereby suppressing cytokine production [42] However, HDACinhibition can also lead to the induction of Bim in lymphoid tumor cells [43] and, as shown here, ultimately triggers apoptosis also in Treg cells As noted before, Treg cells were found highly susceptible to apoptosis induction by Fas ligation that may be due to higher Fas surface expression by Treg cells compared to Tcon cells (Ref [44] and Tischner et al., pers observation) Notably, expression of Noxa mRNA was found also to be much higher in splenic Treg cells, as opposed to Tcon cells Since Noxa expression can be induced by antigenic stimulation in peripheral CD8ỵ T cells [45], we assume that stimulation by self-peptide/MHC molecules may contribute to this phenotype but protein levels achieved appear insufficient to cause death However, increased Noxa may prime Treg cells to apoptosis, e.g as it does conventional T cells under glucose limiting conditions [46] Similar to observations made with conventional T cells [8], Treg cell number was augmented in BimÀ/À and vav-Bcl-2 mice Notably, some of these cells developed into double negative Treg cells in the thymus that were then also found in the periphery (Fig 3) Cells in the thymus seem to derive from SP4ỵ Treg cells, as they also stain positive for the TCR-beta chain on their surface (not shown) and presumably escaped negative selection These observations are in line with most recent findings by Zhan et al who provided evidence that these cells are actually autoreactive cells that have been reprogrammed/redirected to commit to an anergic regulatory T cell fate [47] Notably, BimÀ/À and vav-Bcl-2 Treg cells showed reduced expression of Foxp3, CD25, GITR and CTLA-4, the latter only in the thymus (Fig 3), suggesting that these cells may not be fully functional which was confirmed in vitro (Fig 4) and in vivo (Fig 5) While we, and others [32], failed to note substantial differences in genotype matched suppression assays, arguing against the cytokine consumption hypothesis for T cell suppression [18], we also realized that BimÀ/À and vav-Bcl-2 Treg cells displayed a reduced suppressive capacity towards Tcon cells from wild type mice (Fig 4) However, attempts to define differences in calcium signalling, known to be modulated by Bcl-2 [48], in Treg cells were unsuccessful, questioning an intrinsic defect Yet, vav-Bcl-2 Treg cells showed also a strongly impaired suppressive capacity in vivo (Fig 5) However, isolated colitogenic vav-Bcl-2 transgenic Treg cells from these mice showed even higher levels of CD25 and Foxp3 expression, when compared to wild type cells (Fig 6) and were equally potent in suppressing Tcon cells in vitro (Suppl Fig 4C) As this phenomenon coincided with increased total splenocyte number and IFN-g production by Tcon cells in the mesenteric lymph nodes (Fig 5), we assume that the reduced suppressive capacity of Treg cells in vav-Bcl-2 mice is not cell intrinsic and can be overcome during inflammation Consistently, addition of IL-2 in vitro was shown to increase Foxp3 expression in Bim-deficient Treg cells [47], in line with IL-2 signalling impacting directly on Foxp3 promoter activity [49] However, as mice never fully recovered their initial body weight over a period of weeks, we assume that Bcl-2 transgenic Treg cells may be able to recover only transiently or carry additional defects that affect the biological outcome (Fig 5) Another open question is why these “disabled” Treg cells develop in BimÀ/À and vav-Bcl-2 mice in the first place and where they originate Zhan and colleagues propose that these cells accumulate in such large number because they represent survivors of negative selection deviated into an anergic Treg cell fate when mitochondrial cell death is impaired Consistently, these cells more frequently expressed self-reactive TCRs against endogenous superantigens and display low level CD25 expression, presumably as they no longer depend strictly on IL-2 for survival [47] However, our study shows that additional Treg cells lacking CD4 expression accumulate in the thymus (Fig 3C) and that these cells can also arise de novo from CD4ỵ Treg cells during suppression of T cell expansion in vivo (Fig 6) We assume that these cells also represent a pool of cells that should have been deleted, either in the thymus during negative selection, or towards the end of an immune response in the periphery Whether these cells are still functional remains to be determined Conclusions Collectively, our findings demonstrate that cell death along the intrinsic Bcl-2 regulated apoptosis pathway is critical for normal development and function of Treg cells Intrinsic differences in Bcl2 family protein expression impinge on the relative cell death susceptibility of Treg cells to a number of stress signals that may become of pathophysiological importance, such as those elicited by glucocorticoids or cytokine withdrawal Deregulated expression of key effectors of the Bcl-2 family impairs lineage commitment and suppressive capacity but these defects are not imprinted and appear largely reversible during conditions of inflammation Therefore, attempts to increase the yield or lifespan of Treg cells prior therapeutic application by modulation of cell death regulators should be feasible without significantly compromising their regulatory capacity, albeit a certain caveat remains, warranting further detailed investigations Authorship contribution D.T designed and performed most experiments, statistical analysis, prepared figures and wrote paper I.G., I.P., M.K and S.T performed experiments M.D performed histopathological assessment A.V designed research, interpreted data, wrote paper, conceived study G.J.W performed experiments, designed research, interpreted data and wrote paper A.V and G.J.W contributed equally to this study Disclosure of conflicts The authors have no conflicting financial interests Acknowledgments We are grateful to K Rossi and C Soratroi for animal care and technical assistance as well as to G Böck for cell sorting We thank H Acha-Orbea, J Adams and A Strasser for mice and reagents; M Erlacher for help with MLPA analysis This work was supported by grants from the Austrian Science Fund (FWF); SFB021 and START Y212-B12 to A.V.; the Tiroler Krebshilfe to J.G.W and the Daniel Swarovski Fond to D.T Appendix Supplementary material Supplementary data related to this article can be found online at doi:10.1016/j.jaut.2011.12.008 References [1] Strasser A The role of BH3-only proteins in the immune system Nat Rev Immunol 2005;5:189e200 [2] Rieux-Laucat F, Le Deist F, Fischer A Autoimmune lymphoproliferative syndromes: genetic defects of apoptosis pathways Cell Death Diff 2003;10: 124e33 [3] Tischner D, Woess C, Ottina E, Villunger A Bcl-2-regulated cell death signalling in the prevention of autoimmunity Cell Death Dis 2010;1:e48 D Tischner et al / Journal of Autoimmunity 38 (2012) 59e69 [4] Strasser A, Harris AW, Bath ML, Cory S Novel primitive lymphoid tumours induced in transgenic mice by cooperation between myc and bcl-2 Nature 1990;348:331e3 [5] Campbell KJ, Bath ML, Turner ML, Vandenberg CJ, Bouillet P, Metcalf D, et al Elevated Mcl-1 perturbs lymphopoiesis, promotes transformation of hematopoietic stem/progenitor cells and enhances drug-resistance Blood; 2010 [6] Egle A, Harris AW, Bath ML, O’Reilly L, Cory S VavP-Bcl2 transgenic mice develop follicular lymphoma preceded by germinal center hyperplasia Blood 2004;103:2276e83 [7] Strasser A, Whittingham S, Vaux DL, Bath ML, Adams JM, Cory S, et al Enforced BCL2 expression in B-lymphoid cells prolongs antibody responses and elicits autoimmune disease Proc Natl Acad Sci U S A 1991;88:8661e5 [8] Bouillet P, Metcalf D, Huang DCS, Tarlinton DM, Kay TWH, Köntgen F, et al Proapoptotic Bcl-2 relative Bim required for certain apoptotic responses, leukocyte homeostasis, and to preclude autoimmunity Science (New York, N.Y.) 1999;286:1735e8 [9] Hutcheson J, Scatizzi JC, Siddiqui AM, Haines 3rd GK, Wu T, Li QZ, et al Combined deficiency of proapoptotic regulators Bim and Fas results in the early onset of systemic autoimmunity Immunity 2008;28:206e17 [10] Doreau A, Belot A, Bastid J, Riche B, Trescol-Biemont MC, Ranchin B, et al Interleukin 17 acts in synergy with B cell-activating factor to influence B cell biology and the pathophysiology of systemic lupus erythematosus Nat Immunol 2009;10:778e85 [11] Mehrian R, Quismorio Jr FP, Strassmann G, Stimmler MM, Horwitz DA, Kitridou RC, et al Synergistic effect between IL-10 and bcl-2 genotypes in determining susceptibility to systemic lupus erythematosus Arthritis Rheum 1998;41:596e602 [12] Lu LF, Rudensky A Molecular orchestration of differentiation and function of regulatory T cells Genes Dev 2009;23:1270e82 [13] Mucida D, Pino-Lagos K, Kim G, Nowak E, Benson MJ, Kronenberg M, et al Retinoic acid can directly promote TGF-beta-mediated Foxp3(ỵ) Treg cell conversion of naive T cells Immunity 2009;30:471e2 [author reply 472e473] [14] Vignali DA, Collison LW, Workman CJ How regulatory T cells work Nat Rev Immunol 2008;8:523e32 [15] Wing K, Sakaguchi S Regulatory T cells exert checks and balances on self tolerance and autoimmunity Nat Immunol 2010;11:7e13 [16] Ren X, Ye F, Jiang Z, Chu Y, Xiong S, Wang Y Involvement of cellular death in TRAIL/DR5-dependent suppression induced by CD4(ỵ)CD25(ỵ) regulatory T cells Cell Death Diff 2007;14:2076e84 [17] Strauss L, Bergmann C, Whiteside TL Human circulating CD4ỵCD25highFoxp3ỵ regulatory T cells kill autologous CD8ỵ but not CD4ỵ responder cells by Fas-mediated apoptosis J Immunol 2009;182:1469e80 [18] Pandiyan P, Zheng L, Ishihara S, Reed J, Lenardo MJ CD4ỵCD25ỵFoxp3ỵ regulatory T cells induce cytokine deprivation-mediated apoptosis of effector CD4ỵ T cells Nat Immunol 2007;8:1353e62 [19] Gambineri E, Torgerson TR, Ochs HD Immune dysregulation, polyendocrinopathy, enteropathy, and X-linked inheritance (IPEX), a syndrome of systemic autoimmunity caused by mutations of FOXP3, a critical regulator of T-cell homeostasis Curr Opin Rheumatol 2003;15:430e5 [20] Brunkow ME, Jeffery EW, Hjerrild KA, Paeper B, Clark LB, Yasayko SA, et al Disruption of a new forkhead/winged-helix protein, scurfin, results in the fatal lymphoproliferative disorder of the scurfy mouse Nat Genet 2001;27:68e73 [21] Lahl K, Loddenkemper C, Drouin C, Freyer J, Arnason J, Eberl G, et al Selective depletion of Foxp3ỵ regulatory T cells induces a scurfy-like disease J Exp Med 2007;204:57e63 [22] Huan J, Culbertson N, Spencer L, Bartholomew R, Burrows GG, Chou YK, et al Decreased FOXP3 levels in multiple sclerosis patients J Neurosci Res 2005;81: 45e52 [23] Kumar M, Putzki N, Limmroth V, Remus R, Lindemann M, Knop D, et al CD4ỵCD25ỵFoxP3ỵ T lymphocytes fail to suppress myelin basic proteininduced proliferation in patients with multiple sclerosis J Neuroimmunol 2006;180:178e84 [24] Haribhai D, Lin W, Relland LM, Truong N, Williams CB, Chatila TA Regulatory T cells dynamically control the primary immune response to foreign antigen J Immunol 2007;178:2961e72 [25] Ogilvy S, Metcalf D, Print CG, Bath ML, Harris AW, Adams JM Constitutive bcl2 expression throughout the hematopoietic compartment affects multiple lineages and enhances progenitor cell survival Proc Natl Acad Sci U S A 1999; 96:14943e8 [26] Ranger AM, Zha J, Harada H, Datta SR, Danial NN, Gilmore AP, et al Baddeficient mice develop diffuse large B cell lymphoma Proc Natl Acad Sci U S A 2003;100:9324e9 69 [27] Labi V, Erlacher M, Krumschnabel G, Manzl C, Tzankov A, Pinon J, et al Apoptosis of leukocytes triggered by acute DNA damage promotes lymphoma formation Genes Dev 2010;24:1602e7 [28] Villunger A, Michalak EM, Coultas L, Mullauer F, Bock G, Ausserlechner MJ, et al p53- and drug-induced apoptotic responses mediated by BH3-only proteins puma and noxa Science (New York, N.Y.) 2003;302:1036e8 [29] Eldering E, Spek CA, Aberson HL, Grummels A, Derks IA, de Vos AF, et al Expression profiling via novel multiplex assay allows rapid assessment of gene regulation in defined signalling pathways Nucleic Acids Res 2003;31:e153 [30] Fontenot JD, Rasmussen JP, Gavin MA, Rudensky AY A function for interleukin in Foxp3-expressing regulatory T cells Nat Immunol 2005;6:1142e51 [31] Fontenot JD, Rasmussen JP, Williams LM, Dooley JL, Farr AG, Rudensky AY Regulatory T cell lineage specification by the forkhead transcription factor foxp3 Immunity 2005;22:329e41 [32] Szymczak-Workman AL, Delgoffe GM, Green DR, Vignali DA Cutting edge: regulatory T cells not mediate suppression via programmed cell death pathways J Immunol 2011;187:4416e20 [33] Ludwinski MW, Sun J, Hilliard B, Gong S, Xue F, Carmody RJ, et al Critical roles of Bim in T cell activation and T cell-mediated autoimmune inflammation in mice J Clin Investig 2009;119:1706e13 [34] Opferman JT, Letai A, Beard C, Sorcinelli MD, Ong CC, Korsmeyer SJ Development and maintenance of B and T lymphocytes requires antiapoptotic MCL1 Nature 2003;426:671e6 [35] Mandal M, Borowski C, Palomero T, Ferrando AA, Oberdoerffer P, Meng F, et al The BCL2A1 gene as a pre-T cell receptor-induced regulator of thymocyte survival J Exp Med 2005;201:603e14 [36] Ma A, Pena JC, Chang B, Margosian E, Davidson L, Alt FW, et al Bclx regulates the survival of double-positive thymocytes Proc Natl Acad Sci U S A 1995;92: 4763e7 [37] Bouillet P, Purton JF, Godfrey DI, Zhang L-C, Coultas L, Puthalakath H, et al BH3-only Bcl-2 family member Bim is required for apoptosis of autoreactive thymocytes Nature 2002;415:922e6 [38] Aschenbrenner K, D’Cruz LM, Vollmann EH, Hinterberger M, Emmerich J, Swee LK, et al Selection of Foxp3ỵ regulatory T cells specific for self antigen expressed and presented by Aireỵ medullary thymic epithelial cells Nat Immunol 2007;8:351e8 [39] Moon JJ, Dash P, Oguin 3rd TH, McClaren JL, Chu HH, Thomas PG, et al Quantitative impact of thymic selection on Foxp3ỵ and Foxp3 subsets of self-peptide/MHC class II-specic CD4ỵ T cells Proc Natl Acad Sci U S A 2011; 108:14602e7 [40] Chen X, Murakami T, Oppenheim JJ, Howard OM Differential response of murine CD4ỵCD25ỵ and CD4ỵCD25 T cells to dexamethasone-induced cell death Eur J Immunol 2004;34:859e69 [41] Guimond M, Veenstra RG, Grindler DJ, Zhang H, Cui Y, Murphy RD, et al Interleukin signaling in dendritic cells regulates the homeostatic proliferation and niche size of CD4ỵ T cells Nat Immunol 2009;10:149e57 [42] Tao R, de Zoeten EF, Ozkaynak E, Chen C, Wang L, Porrett PM, et al Deacetylase inhibition promotes the generation and function of regulatory T cells Nat Med 2007;13:1299e307 [43] Lindemann RK, Newbold A, Whitecross KF, Cluse LA, Frew AJ, Ellis L, et al Analysis of the apoptotic and therapeutic activities of histone deacetylase inhibitors by using a mouse model of B cell lymphoma Proc Natl Acad Sci U S A 2007;104:8071e6 [44] Weiss EM, Schmidt A, Vobis D, Garbi N, Lahl K, Mayer CT, et al Foxp3mediated suppression of CD95L expression confers resistance to activationinduced cell death in regulatory T cells J Immunol 2011;187:1684e91 [45] Wensveen FM, van Gisbergen KP, Derks IA, Gerlach C, Schumacher TN, van Lier RA, et al Apoptosis threshold set by Noxa and Mcl-1 after T cell activation regulates competitive selection of high-affinity clones Immunity 2010;32: 754e65 [46] Alves NL, Derks IA, Berk E, Spijker R, van Lier RA, Eldering E The Noxa/Mcl-1 axis regulates susceptibility to apoptosis under glucose limitation in dividing T cells Immunity 2006;24:703e16 [47] Zhan Y, Zhang Y, Gray D, Carrington EM, Bouillet P, Ko HJ, et al Defects in the Bcl-2-regulated apoptotic pathway lead to preferential increase of CD25 low Foxp3ỵ anergic CD4ỵ T cells J Immunol 2011;187:1566e77 [48] Rong Y, Distelhorst CW Bcl-2 protein family members: versatile regulators of calcium signaling in cell survival and apoptosis Annu Rev Physiol 2008;70: 73e91 [49] Zorn E, Nelson EA, Mohseni M, Porcheray F, Kim H, Litsa D, et al IL-2 regulates FOXP3 expression in human CD4ỵCD25ỵ regulatory T cells through a STATdependent mechanism and induces the expansion of these cells in vivo Blood 2006;108:1571e9  ... our findings demonstrate that cell death along the intrinsic Bcl- 2 regulated apoptosis pathway is critical for normal development and function of Treg cells Intrinsic differences in Bcl2 family... of apoptosis pathways Cell Death Diff 20 03;10: 124 e33 [3] Tischner D, Woess C, Ottina E, Villunger A Bcl- 2- regulated cell death signalling in the prevention of autoimmunity Cell Death Dis 20 10;1:e48... compared to their BimÀ/À and vav -Bcl- 2 counterparts of Treg cell markers in foxp3gfpwild type, foxp3gfpBimÀ/À and foxp3gfpvav -Bcl- 2 mice, the latter two showing increased Treg cell numbers While in wild