Foxp3-Deficient Regulatory T Cells Do Not Revert into Conventional Effector CD4 + T Cells but Constitute a Unique Cell Subset

This information is current as Michal Kuczma, Robert Podolsky, Nikhil Garge, Danielle of September 26, 2021. Daniely, Rafal Pacholczyk, Leszek Ignatowicz and Piotr Kraj J Immunol 2009; 183:3731-3741; Prepublished online 26 August 2009;

doi: 10.4049/jimmunol.0800601 Downloaded from http://www.jimmunol.org/content/183/6/3731

Supplementary http://www.jimmunol.org/content/suppl/2009/08/26/jimmunol.080060 Material 1.DC1 http://www.jimmunol.org/ References This article cites 50 articles, 18 of which you can access for free at: http://www.jimmunol.org/content/183/6/3731.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 26, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2009 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Foxp3-Deficient Regulatory T Cells Do Not Revert into Conventional Effector CD4؉ T Cells but Constitute a Unique Cell Subset1,2

Michal Kuczma, Robert Podolsky, Nikhil Garge, Danielle Daniely, Rafal Pacholczyk, Leszek Ignatowicz, and Piotr Kraj3

؉ Homeostasis in the immune system is maintained by specialized regulatory CD4 T cells (Treg) expressing transcription factor Foxp3. According to the current paradigm, high-affinity interactions between TCRs and class II MHC-peptide complexes in thymus “instruct” developing thymocytes to up-regulate Foxp3 and become Treg cells. However, the loss or down-regulation of

Foxp3 does not disrupt the development of Treg cells but abrogates their suppressor function. In this study, we show that

Foxp3-deficient Treg cells in scurfy mice harboring a null mutation of the Foxp3 gene retained cellular features of Treg cells Downloaded from including in vitro anergy, impaired production of inflammatory cytokines, and dependence on exogenous IL-2 for proliferation and homeostatic expansion. Foxp3-deficient Treg cells expressed a low level of activation markers, did not expand relative to other CD4؉ T cells, and produced IL-4 and immunomodulatory cytokines IL-10 and TGF-␤ when stimulated. Global gene expression profiling revealed significant similarities between Treg cells expressing and lacking Foxp3. These results argue that Foxp3 defi- ؉ ciency alone does not convert Treg cells into conventional effector CD4 T cells but rather these cells constitute a distinct cell subset with unique features. The Journal of Immunology, 2009, 183: 3731–3741. http://www.jimmunol.org/

4 atural regulatory T (Treg) cells are produced in the thymus Decreased function of Treg cells has been associated with vari- where they initiate expression of the X chromosome-linked ous autoimmune disorders in humans and mice (9). The reduced transcription factor Foxp3 which endows these cells with level of Foxp3 expression correlated with impaired T function N reg suppressor function (1–3). Recognition of class II MHC loaded with and was found in such autoimmune diseases as myasthenia gravis agonist peptide by the developing thymocytes augmented the gener- and multiple sclerosis (10, 11). The most conspicuous deficiency ation of T cells specific for cognate Ag. This led to the hypothesis reg of Treg function is observed in the human autoimmune disease that Foxp3 expression and selection of Treg cells is “instructed” by IPEX (immune dysregulation, polyendocrinopathy, enteropathy, by guest on September 26, 2021 high-affinity interaction between TCR and peptide-MHC complexes X-linked) and the corresponding disease in scurfy mice (12, 13). and further implied that Treg cells express TCRs with higher affinity Affected males suffer from fatal, multiorgan, lymphoproliferative for self-Ags than conventional T cells (4). Foxp3 expression was pos- disease mediated by CD4ϩ T cells (14, 15). Mutations in the tulated to decrease sensitivity of TCR stimulation of Treg cells and Foxp3 gene affecting its function were found to be the molecular explained why these cells are anergic in vitro and do not become basis of IPEX and scurfy diseases. pathogenic in vivo despite expressing self-reactive TCRs (5). How- Recent analyses of mice expressing defective alleles of Foxp3 ever, an alternative model, where Foxp3 up-regulation may happen have shown that Foxp3 deficiency does not impair lineage com- regardless of the TCR affinity for the selecting peptide ligand, has mitment and development of T cells (16, 17). Thus, Foxp3 ex- never been disproved and the role of self-reactivity in the develop- reg pression might be a concluding, rather than a causal event in the ment of Treg cells remains controversial (6–8). Treg cell lineage differentiation that endows thymocytes that had

already initiated the transcriptional program of Treg cells with sup- pressor function. Foxp3 binds to regulatory regions of hundreds of Center for Biotechnology and Genomic Medicine, Medical College of Georgia, Au- gusta, GA 30912 genes in Treg cells, many of which control the T cell response to Received for publication February 26, 2008. Accepted for publication July 14, 2009. Ag stimulation (18, 19). The impaired activity of Foxp3 could The costs of publication of this article were defrayed in part by the payment of page result in the abrogation of molecular control mechanisms in Treg charges. This article must therefore be hereby marked advertisement in accordance cells and restoration of CD4ϩ T cell effector functions. Unfortu- with 18 U.S.C. Section 1734 solely to indicate this fact. nately, little is known about the extent of diversity in the level of 1 This work was supported by National Institutes of Health Grant R01 CA107349- Foxp3 expression in the T cells of healthy subjects and how 01A1 and a Georgia Cancer Coalition award. L.I. was supported by basic National reg Institutes of Health Research Grant Al041145-07A1 and the Roche Organ Transplan- Foxp3 down-regulation affects Treg cellular functions. Investigat- tation Research Foundation. ing the properties of Foxp3-deficient Treg cells could not only re- 2 The GeneChip expression data were deposited in the Gene Expression Omnibus veal cellular functions controlled by Foxp3 but also help better database http://www.ncbi.nlm.nih.gov/geo/. The accession number is GSE11775. assess the potential of immunotherapy aimed at modulating Foxp3 3 Address correspondence and reprint requests to Dr. Piotr Kraj, Center for Biotech- expression. Since Treg cells may constitute a reservoir of self-re- nology and Genomic Medicine, Medical College of Georgia, CA-4141, Augusta, GA ϩ 30912. E-mail address: [email protected] active CD4 T cells, uncovering the consequences of Foxp3 4 Abbreviations used in this paper: Treg, regulatory T; IPEX, immune dysregulation, down-regulation could explain the pathogenesis of multiple auto- polyendocrinopathy, enteropathy, X-linked; BAC, bacterial artificial chromosome; immune diseases, in particular, the contribution of Foxp3-deficient galK, galactokinase; Teff, effector T. ϩ Treg cells to autoimmune pathology. CD4 T cells expressing mu- Copyright © 2009 by The American Association of Immunologists, Inc. 0022-1767/09/$2.00 tant forms of Foxp3 were found in IPEX patients but their role in www.jimmunol.org/cgi/doi/10.4049/jimmunol.0800601 3732 Foxp3-DEFICIENT Treg CELLS autoimmune pathology remains unknown (20–22). These cells closed circular BAC DNA into oocytes from C57BL/6 mice. Founders could represent thymocytes that attempted T cell development were genotyped by PCR with primers specific to GFP (forward primer, reg Ј Ј Ј and migrated to the periphery but retained at least some properties 5 -GTGCCCATCCTGGTCGAGCTGGACGG3- and reverse primer, 5 - CTTTGCTCAGGGCGGACTGGGTGCTCAGG-3Ј). Five founders ex- of functional Treg cells despite losing suppressor function. Alter- pressed the transgene and transmitted it to progeny. Transgenic founder 90 natively, these cells could represent aggressive, self-reactive T was selected for further crossing. cells that originate from the T lineage and significantly contrib- reg Mice ute to the severity of IPEX disease by producing IL-2 and IFN-␥ (22). Since conventional human CD4ϩ T cells transiently up-reg- Scurfy and C57BL/6 mice were purchased from The Jackson Laboratory GFP ulate Foxp3 upon activation, it was not possible to determine the and crossed with transgenic Foxp3 mice. Mice were housed under spe- developmental origin of these cells (23). cific pathogen-free conditions and used according to the guidelines of the Animal Care and Use Committee of the Medical College of Georgia. We have established that Foxp3-deficient Treg cells in sick scurfy males, in the absence of functional Treg cells, remained Flow cytometry and cell sorting ϩ quiescent, did not expand relative to other CD4 T cells, and ex- Single-cell suspensions were prepared from thymi and lymph nodes by pressed a lower level of activation markers compared with effector mechanical disruption and cells were stained with Abs available com- CD4ϩ T cells. In in vitro assays, SfFoxp3GFPϩ cells did not pro- mercially (eBioscience or BD Biosciences). Cells were analyzed using duce IL-2 and poorly responded to TCR stimulation. Moreover, a FACSCanto flow cytometer (BD Biosciences) and FACSDiva or WinList GFPϩ software. Cells were sorted on a MoFlo cell sorter (DakoCytomation). For SfFoxp3 cells produced much fewer inflammatory cyto- ϩ ϩ some experiments, CD4 T cells were negatively sorted using commercial kines than effector CD4 T cells, except IL-4, but were able to kit and an AutoMACS magnetic cell sorter (Miltenyi Biotec). Intracellular produce IL-10 and TGF-␤. Gene expression analysis showed that staining for Foxp3 and Ki-67 was performed according to the manufactur- Downloaded from ers’ instructions (eBioscience and BD Biosciences, respectively). transcription of many Treg-specific genes is similar in GFPϩ SfFoxp3 and functional Treg cells. By comparing gene ex- Proliferation assay pression in effector and T cells isolated from scurfy and healthy reg ϫ 4 mice, we defined T -specific, Foxp3-independent gene signature. Lymph node proliferation assays were performed with 3–5 10 cells reg isolated from Foxp3GFP or SfFoxp3GFP mice. Cells were sorted directly Analysis of T cell hybridomas derived from effector and onto 96-well plates using a MoFlo sorter and cultured for 3 days. Wells GFPϩ SfFoxp3 cells revealed that the frequency of self-reactive were coated overnight with anti-CD3 (10 ␮g/ml) and anti-CD28 (1 ␮g/ml) http://www.jimmunol.org/ TCRs is similar in both cell subsets. In conclusion, Foxp3 defi- Abs. Proliferation responses were measured by adding 1 ␮Ci/well of [3H]thymidine on day 3 of a 4-day culture. ciency does not convert Treg cells into conventional, self-reactive GFPϩ effector cells and SfFoxp3 cells retain cellular features of Treg Inhibition assay cells in inflammatory environments. Despite poor potential for clonal ϩ GFPϪ ϫ 4 expansion, Foxp3-deficient T cells may modulate immune re- Sorted CD4 Foxp3 cells (5 10 /well) were incubated on a 96-well reg plate with irradiated splenocytes (5 ϫ 104/well, 3000 rad) and soluble sponses by secreted cytokines, especially IL-4, to shift the immune anti-CD3 (5 ␮g/ml). Various numbers of sorted CD4ϩFoxp3GFPϩ cells response toward Th2 effector cells. At the same time, conventional (1–5 ϫ 104/well) were added. Cells were sorted using a MoFlo sorter. After CD4ϩ T cells express a highly activated phenotype and produce a 3 days culture proliferation was measured by adding 1 ␮Ci/well of 3 very high level of IL-2 and inflammatory cytokines. Our data suggest [ H]thymidine. by guest on September 26, 2021 that the vast majority of T cells that cause autoimmune pathology in Adoptive transfer scurfy mice originate from conventional CD4ϩ T cells. Donor cells for adoptive transfer were isolated by flow cytometry sorting of total lymph node CD4ϩ T cells or CD4ϩGFPϩ or CD4ϩGFPϪ cell Materials and Methods subsets from SfFoxp3GFP or Foxp3GFP mice. The number of cells indicated Foxp3-GFP-transgenic construct in each experiment was transferred i.v. into recipient TCR␣ chain knockout or scurfy mice. In cotransfer experiments, congenic C57BL/6 or scurfy A bacterial artificial chromosome (BAC) clone (RP23-446O15, 186.8 kb) mice expressing different alleles of CD45 (Ly5) were used as cell donors. isolated from a C57BL/6 genomic library and consisting of the Foxp3 gene Recipient mice were analyzed 4–5 wk after adoptive transfer into lym- was purchased from BacPac. Three other known genes are located on the phopenic mice and 10 days after transfer into scurfy mice. BAC DNA, Ppp1r3f, Ccdc22, and Cacna1f, but none of them was reported to be involved in T cell function. GFP followed by polyadenylation signal RT-PCR was introduced in frame with the Foxp3 translation initiation site into exon 1 of the Foxp3 gene using the BAC recombineering system (24). BAC RNA was isolated from sorted cells (104 cells/sample) with a RNeasy Mini modifications were done in the Escherichia coli strain SW102 by a two- Kit (Qiagen) and reverse transcribed using a Superscript kit (Invitrogen) step recombination process with galactokinase (galK) as the selection gene. according to the manufacturers’ instructions. The quantities of cDNA were In the first step, galK was inserted into the first exon of the Foxp3 gene by normalized for ␤-actin. Foxp3 cDNA was amplified with sense primer homologous recombination with the targeting plasmid consisting of 5Ј arm, 5Ј-ATCCAGCCTGCCTCTGACAAGAACC- 3Ј and reverse primer 5Ј- galK, and 3Ј arm. The galK gene expression was driven by the EM7 pro- GGGTTGTCCAGTGGACGCACTTGGAGC-3Ј. These primers distin- moter. Homologous recombination occurred within the 5Ј and 3Ј arms guish between amplification product of the endogenous Foxp3 gene derived from the Foxp3 gene, and the galK gene was inserted into BAC (401 bp) and the transgenic transcript (1357 bp). DNA. Bacteria containing the recombinant BAC were selected on minimal medium with galactose as the only carbon source. The modified BAC clone Western blotting was subjected to another recombination event with the second targeting ϩ Ϫ Foxp3 protein was detected in sorted (105 cells/sample) CD4 Foxp3GFP construct consisting of the 5Ј arm, the GFP-poly(A) cassette, and the 3Ј ϩ ϩ and CD4 Foxp3GFP cells. Cells were lysed in the gel-loading buffer and arm. Second-step recombinants were selected against galK on minimal resolved on a 10% polyacrylamide gel. Proteins were transferred onto medium plates with glycerol as the carbon source and 2-deoxygalactose as PVDF membrane (Millipore). Membranes were probed with anti-Foxp3 the selecting agent. 2-deoxygalactose is phosphorylated to toxic 2-deox- Ab eBio7979 (eBioscience) followed by goat anti-mouse polyclonal Ab ygalactose-1-phosphate by bacteria expressing galK so only bacteria that coupled with HRP (Bio-Rad). Membranes were developed with an ECL kit replaced galK with GFP survive. (Amersham Biosciences) according to the manufacturer’s instructions. The design of the GFP expression cassette ensures that Foxp3 transcript, initiated at exon Ϫ2b, ϳ6.1 kb upstream from the first coding exon, is Cytokine detection cleaved downstream of GFP and polyadenylated. The remaining fragment of the transcript is degraded in the nucleus since it cannot be capped with Production of cytokines by CD4ϩFoxp3GFPϪ and CD4ϩFoxp3GFPϩ T cells methylguanosine and transported to the cytoplasm. This ensures that tran- was assessed using a Q-Plex mouse cytokine array (Quansys Biosciences). scripts originating from the transgene cannot be translated into functional Cells were sorted onto 96-well plates (5 ϫ 104/well) coated with anti-CD3 Foxp3 protein. Transgenic mice were produced by pronuclear injection of (10 ␮g/ml) and anti-CD28 (1 ␮g/ml) Abs. After 30 h, the supernatant was The Journal of Immunology 3733 collected and used to measure cytokine levels according to the manufac- hypothesis. The advantage of the B statistic is that it accurately ranks the genes turer’s instruction. The chemiluminescence image was acquired with a Fu- in order of likelihood of being differentially expressed. Choosing a cutoff for jifilm LAS-3000 imaging system. Alternatively, cytokine levels were mea- B, however, is just as challenging as using any other statistic. We called all sured by ELISA using commercial kits according to manufacturer’s genes with a B Ն1.5 as significant, since the evidence for the alternative would instructions (eBioscience). For ELISA, 2 ϫ 105 cells were stimulated with no longer be considered weak. This choice of cutoff also seemed reasonable anti-CD3 and anti-CD28 Abs and supernatants were collected after 30 h. since the q values (expected false discovery rates) for those genes we called Cytokine transcripts were detected in sorted SfFoxp3GFPϪ and significant were ϳ0.01. SfFoxp3GFPϩ cells without in vitro stimulation by real-time PCR. cDNA was produced as described above. The quantities of cDNA were nor- Results ␤ ␤ Ј malized for -actin. -Actin was amplified with the sense primer 5 - Expression of GFP reporter defines a population of regulatory CCTTCTACAATGAGCTGCGTGTGGC-3Ј and antisense primer 5Ј-CA ϩ TGAGGTAGTCTGTCAGGTCC-3Ј. Cytokine cDNA was amplified CD4 T cells in healthy and Foxp3-deficient scurfy mice with the following primers: IL-2 sense, 5Ј-CCTTGCTAATCACTCC To facilitate analysis, Foxp3-deficient T cells in scurfy males Ј Ј Ј reg TCACA-3 and antisense, 5 -GAGCTCCTGTAGGTCCATCA-3 ; IL-4 were identified by expressing GFP reporter. Transgenic mice ex- sense, 5Ј-CAAGGTGCTTCGCATATTTT-3Ј and antisense, 5Ј-ATCC ATTTGCATGATGCTCT-3Ј; IL-10 sense, 5Ј-AGTGGAGCAGGTGA pressing GFP controlled by the Foxp3 regulatory sequences AGAGTG-3Ј and antisense, 5Ј-TTCGGAGAGAGGTACAAACG-3Ј; (Foxp3GFP mice) and scurfy mice were crossed (to produce IL-17 sense, 5Ј-AGGCCCTCAGACTACCTCAA-3Ј and antisense, 5Ј- SfFoxp3GFP mice) and males coexpressing the mutant Foxp3 allele Ј ␥ Ј CAGGATCTCTTGCTGGATGA-3 ; IFN- sense, 5 -AGTGGAGCAG and Foxp3GFP reporter transgene, not located on X chromosome, GTGAAGAGTG-3Ј and antisense, 5Ј-TTCGGAGAGAGGTACAAA CG-3Ј; and TGF-␤ sense, 5Ј-GCTACCATGCCAACTTCTGT-3Ј and were examined. To ensure cell-type-specific expression of a re- antisense, 5Ј-CGTAGTAGACGATGGGCAGT-3Ј. cDNA prepared porter gene, a BAC clone encompassing the whole Foxp3 tran- from cells known to produce a particular cytokine was used as a pos- scription unit was modified by inserting a reporter cassette encod- Downloaded from itive control. For IL-2, IFN-␥, IL-4, and IL-17 cDNA from cells stim- ing GFP followed by the STOP codon and the poly(A) signal ulated in vitro under neutral, Th2, or Th17 conditions was used. Control ␤ sequence into exon 1 in frame with the start codon of the Foxp3 cDNA for IL-10 and TGF- was prepared from Treg cells. Quantitative PCR was performed on the Bio-Rad iCycler using SYBR Green detec- gene (Fig. 1). The design of the expression cassette prevents over- tion. The PCR conditions were: denaturation at 95oC for 2 min. fol- expression of the Foxp3 from the BAC transgene, which is known lowed by 50 cycles of denaturation at 95oC for 20 s, annealing at 58oC to alter the function of CD4ϩ T cells, and production of the Foxp3- for 10 s, and elongation at 72oC for 20 s. The relative levels of cytokine GFP GFP fusion protein or truncated Foxp3 (28). Transgenic Foxp3 http://www.jimmunol.org/ mRNA were determined. For each sample, the per cell mRNA level of GFP the cytokine gene was determined by normalizing the experimentally and C57BL/6 mice as well as scurfy and SfFoxp3 mice, respec- determined mRNA level of the cytokine gene of interest to internal tively, had equivalent numbers, percentages, and cell surface phe- control ␤-actin mRNA level. The mRNA level of the cytokine gene of notypes of all T and non-T cell subsets, including CD4ϩCD25ϩ T interest in a sample stimulated in neutral, Th2, and Th17 conditions was cells, in thymus, lymph nodes, and spleen (Fig. 2, A and B, and arbitrarily set as 1. The relative level of the mRNA level of the cytokine data not shown). Approximately 90% of CD4ϩCD25ϩ T cells in gene of interest was calculated and presented in a bar graph. healthy mice expressed Foxp3 (29, 30). Flow cytometry, RT-PCR, ϩ Production and analysis of CD4 hybridomas and Western blot analyses show that only GFPϩ, and not GFPϪ, ϩ T cells hybridomas were produced from flow cytometer-sorted SfFoxp3GFPϪ CD4 lymphocytes expressed Foxp3, demonstrating reliable ex- GFPϩ GFP by guest on September 26, 2021 and SfFoxp3 cells directly fused to BW thymoma deficient in endog- pression of the Foxp3 transgene in Treg cells or organs known enous ␣␤ TCR as described previously (25). Hybridomas were tested for to contain CD4ϩ T cells (Fig. 2, C, E, and F). SfFoxp3GFPϩ males reactivity to self-Ags using a standard IL-2 release assay (8). Hybridomas developed lymphoproliferative disease indistinguishable from the produced in two independent experiments (42 and 44 hybridomas from Ϫ ϩ scurfy ϳ SfFoxp3GFP cells and 9 and 25 hybridomas from SfFoxp3GFP cells) ex- disease in nontransgenic males and died at 3–4 wk of age. pressed TCR and CD4 and responded to stimulation with plate-bound Abs Mice displayed morphological symptoms of scurfy disease includ- by producing IL-2. To determine hybridoma reactivity independent of IL-2 ing runting, splenomegaly, dermatitis, malformed ears, and greatly production we assessed CD69 expression on hybridomas stimulated with enlarged lymph nodes (data not shown) (13). Heterozygous scurfy plate-bound Abs or autologous splenocytes. The results of both assays were females remained healthy and the phenotype of T cell populations consistent. The fraction of self-reactive hybridomas was calculated by di- GFP viding the number of hybridomas responding to splenocytes by the number was the same as in Foxp3 mice (data not shown). In vitro GFP of hybridomas responding to stimulation by plate-bound Abs. expression of the Foxp3 reporter remained stable for Ͼ2 days in Ag-stimulated GFPϩ cells from healthy and scurfy mice and Microarray analysis was further inducible by TGF-␤ treatment (data not shown). In GFP RNA was prepared from sorted cell subsets using a RNeasy kit (Qiagen). Treg conclusion, the Foxp3 reporter transgene reliably defines the ϩ and conventional CD4 T cells from SfFoxp3GFP and Foxp3GFP mice were population of T cells. analyzed in triplicates. RNA was amplified using a TargetAmp kit (Epicentre). reg The resulting cRNA was hybridized to Affymetrix GeneChip M430 2.0 Plus. The loss of suppressor function by SfFoxp3GFPϩCD4ϩ T cells is Microarray data were first normalized using RMA and subsequently not accompanied by the acquisition of a high level of activation analyzed using Linear Models for Microarray Data (26, 27). We analyzed all arrays as a factorial experiment in which strain (Sf vs wild type) was one markers ϩ GFPϩ GFPϩ factor and cell type (Treg vs effector T (Teff)) was a second factor, along CD4 Foxp3 but not SfFoxp3 T cells inhibited activation with the interaction of strain and cell type. Genes whose response was ϩ of CD4 T cells in vitro (Fig. 2D). Adoptively transferred, purified Foxp3 dependent were those found significant for the interaction, regard- ϩ GFP GFP less of significance for the main effects. Genes with no significant inter- CD4 T cells from SfFoxp3 , but not Foxp3 , mice induced action and no significant response to Foxp3, but having a significant dif- autoimmune disease in T cell-deficient recipients, demonstrating that ference between strains, are those genes that are strain specific regardless ϩ CD4 T cells in the former mice lacked functional Treg cells (data not of Foxp3 expression. Genes with no significant interaction and no signif- shown). Thus, the in vitro assay and cell transfer studies demonstrate icant difference between strains, but having a significant difference between GFPϩ ϩ T and T cells are those genes that respond to Foxp3 expression equally the lack of suppressor function of SfFoxp3 CD4 T cells. reg eff GFPϩ in both strains, with no differences between strains. The advantage to Li- Foxp3 T cells were preferentially found in the population near Models for Microarray Data is that the B statistic (log posterior odds of activated CD44ϩCD62LϪ cells, consistent with the fact that of differential expression) used in this analysis quantifies the evidence for Treg cells express higher levels of CD44 and lower levels of the alternate hypothesis vs the evidence for the null hypothesis. Since B is ϩ CD62L than conventional CD4 T cells (Fig. 3A). A large pro- on a log scale,aBof0indicates that both the alternate and null hypotheses ϩ are equally likely. If the B statistic is positive, then the evidence supports the portion of CD4 T cells in scurfy mice exhibit an activated alternative hypothesis of some difference, while a negative B supports the null phenotype consistent with severity of the disease. Surprisingly, the 3734 Foxp3-DEFICIENT Treg CELLS

CD62L expression on SfFoxp3GFPϩ and conventional CD4ϩ T cells (Fig. 3B). SfFoxp3GFPϩ cells only modestly up-regulated CD44 and down-regulated CD62L compared with the conven- tional CD4ϩ T cells, suggesting that they are less responsive to activation by self-Ags. In summary, expression of activation mark- ers suggests that effector CD4ϩ T cells are more sensitive to ac-

tivation by self-Ags then Treg cells that lose Foxp3 expression.

The expansion and phenotype stability of SfFoxp3GFPϩ T cells The proportion of Foxp3-expressing cells was similar in the lymph nodes of Foxp3GFP and SfFoxp3GFP-transgenic mice and did not increase with disease progression, suggesting that these cells do not have a proliferative advantage over effector CD4ϩ T cells (Fig. 3, C and D). Tissue infiltrates in peripheral organs were dominated by SfFoxp3GFPϪ cells, demonstrating that SfFoxp3GFPϩ cells did not migrate and selectively accumulate in peripheral organs af- fected by the autoimmune disease (Fig. 4 and data not shown). ϩ Ϫ To compare the proliferative potential of GFP and GFP cells, Downloaded from the respective cell subsets were sorted and stained with Ki-67- specific Ab. Ki-67 Ag is expressed in all phases of the cell cycle

except G0; therefore, its detection provides an estimate of the frac- tion of dividing cells in a cell population (31). This analysis

showed that in healthy mice the fraction of proliferating Treg cells ϩ

is about twice as large as the fraction of effector CD4 cells and http://www.jimmunol.org/ is consistent with studies showing that steady-state BrdU in- corporation in CD4ϩCD25ϩ cells is higher than in CD4ϩCD25Ϫ cells (Fig. 5A) (32). In contrast, fractions of SfFoxp3GFPϩ and SfFoxp3GFPϪ cells expressing Ki-67 were similar in scurfy mice. Thus, the major relative increase in the fraction of proliferating cells occurred in the SfFoxp3GFPϪ population. These findings im- plied that T cells recruited to the proliferating population prefer- entially originate from effector SfFoxp3GFPϪ T cells. To further evaluate proliferative capacity of SfFoxp3GFPϩ and by guest on September 26, 2021 SfFoxp3GFPϪ cells, we examined homeostatic expansion of the respective cell populations upon transfer into lymphopenic hosts. This assay is commonly used to compare the potential for clonal FIGURE 1. BAC genomic fragment of DNA encompassing Foxp3 expansion of two T cell populations and has been used to estimate gene before (A) and after (B and C) modification to introduce GFP the frequency of self-reactive clones in expanded populations (5). reporter. A, Genes located on the DNA fragment (between vertical ar- ϩ rows) used to produce transgenic mouse. Genes are depicted by green Analysis of the CD4 T cell population in recipient mice shows a GFPϩ rectangles, filled segments represent exons, and empty segments repre- diminished proportion of SfFoxp3 cells and suggests that they GFPϪ sent introns. Direction of transcription is shown for each gene by arrows are outgrown by the effector SfFoxp3 cells (Fig. 5B). In con- ϩ above genes. Ccdc22 gene partially overlaps with untranslated region of trast, the proportion of Foxp3GFP cells in recipient mice recon- Foxp3 and is shown below other genes. B, Recombination strategy used stituted with CD4ϩ cells from healthy mice was similar to the to introduce GFP reporter into exon 1 of the Foxp3 gene. Purple and proportion in the donor population used for transfer. Analysis of pink lines mark DNA segments used as 5Ј and 3Ј arms, respectively, in Ki-67 in transferred cells shows a much larger fraction of prolif- the recombination process. Arrows show location of PCR primers used erating cells in effector SfFoxp3GFPϪ cells than in SfFoxp3GFPϩ to amplify and clone DNA segments of 5Ј and 3Ј arms. Intermediate Ϫ cells as well as Foxp3GFP effector cells from healthy mice (Fig. construct harbors the galK gene driven by the EM7 prokaryotic pro- moter. ATG shows Foxp3 translation initiation codon. The AvrII re- 5C). This is consistent with more efficient expansion of activated donor CD4ϩ T cells from scurfy than from healthy mice. To de- striction enzyme site was used to introduce EM7-galK or GFP into exon ϩ 1 of Foxp3. C, DNA sequence of exon 1 of the Foxp3 gene before termine whether the decreased proportion of GFP T cells in lym- ϩ (upper panel) and after recombination (lower panel). The GFP coding phopenic mice reconstituted with the total population of CD4 T sequence (green letters) was inserted into the AvrII site (red arrow, cells from SfFoxp3GFP mice is caused by down-regulation of the upper panel) in frame with the Foxp3. Intron sequences are shown as GFP reporter upon homeostatic expansion, we have investigated lowercase gray letters and exons are shown in uppercase letters, coding the persistence of SfFoxp3GFPϩ cells in recipient mice (Fig. 5D). fragments are shown in black, and noncoding fragments are shown in To provide exogenous IL-2, we cotransferred SfFoxp3GFPϩ cells gray. A bovine growth hormone poly(A) cassette is shown in orange. with total CD4ϩ T cells from healthy mice into lymphopenic mice. Nucleotides shown in italics were added during the cloning process. Analysis of transferred cells after 4 wk shows that the predominant fraction of SfFoxp3GFPϩ cells retained Foxp3 transcription. In an- other experiment, we were able to detect SfFoxp3GFPϩ cells in population of cells expressing an activated CD44ϩCD62LϪ phe- recipient mice a few months after adaptive transfer. Most of these notype had a smaller fraction of SfFoxp3GFPϩ cells than the pop- cells retained Foxp3GFPϩ expression but the fraction of cells derived Ϫ ϩ ulation of naive CD44 CD62L T cells (data not shown). To from Foxp3-deficient Treg cells in the total population of transferred further compare the surface phenotype, we analyzed CD44 and CD4ϩ T cells declined (data not shown). The Journal of Immunology 3735 Downloaded from http://www.jimmunol.org/

FIGURE 2. Expression of the Foxp3GFP reporter transgene is confined to cells expressing the endogenous Foxp3 gene and does not alter the development of T cells. A and B, Flow cytometry analysis of thymocytes (A) and lymph node cells (B) of 2- to 3-wk-old healthy Foxp3GFP and sick SfFoxp3GFP mice. At least five mice of each kind were analyzed. C, The Foxp3GFP transgene is expressed exclusively in cells expressing endogenous Foxp3. D, CD4ϩGFPϩ Ⅺ ϩ ϩ ࡗ GFP ϩ ϩ Œ GFP ( )orCD4 CD25 ( ) T cells (Treg) isolated from Foxp3 - transgenic mice but not CD4 GFP ( ) cells from SfFoxp3 mice suppress proliferation ϩ Ϫ ϩ of effector CD4 GFP T cells (Teff). E, Foxp3 RNA and protein are expressed only in GFP cells. RT-PCR (upper panel) and Western blot analysis (lower panel) of Foxp3 expression in sorted CD4ϩGFPϩ and CD4ϩGFPϪ subsets from Foxp3GFP (B6) and SfFoxp3GFP (Sf) mice. cDNA was amplified with by guest on September 26, 2021 primers detecting endogenous Foxp3 transcripts. F, Truncated transcripts of the Foxp3GFP transgene extending downstream of the polyadenylation site are absent GFP GFP GFP in SfFoxp3 -transgenic mice. cDNA prepared from Treg cells isolated from SfFoxp3 (mutant, lanes 1and 4) and Foxp3 (wild type (wt), lanes 2 and 5)-transgenic males were amplified with primers spanning the DNA segment of scurfy mutation, in exon 9 of the Foxp3 gene, specific for mutant (lanes 1 and 2) or wild-type transcripts (lanes 4 and 5). The wild-type Foxp3 transcript in scurfy cells could originate only from the Foxp3GFP transgene and is not detected (lane 4). Primer specificity is verified by amplification of mutant and wild-type transcripts with the relevant primers (lanes 1 and 5). PCR was done with the sense primers specific for mutant Foxp3 (5Ј-TCAGGCCTCAATGGACAAAA3-Ј) or wild-type allele (5Ј-CTCAGGCCTCAATGGACAAG-3Ј) and common antisense primer (5Ј-CATCGGATAAGGGTGGCATA-3Ј). Negative control is a PCR without added template (lanes 3 and 6).

It is possible that the contribution of Foxp3-deficient Treg cells to autoimmune pathology was concealed by the loss of the GFP expression. To investigate how stable is the phenotype of SfFoxp3GFPϩ cells in the inflammatory environment, we trans- ferred i.p. equal numbers of CD4ϩLy5.1ϩ/ϩSfFoxp3GFPϩ and CD4ϩLy5.1ϩ/ϪSfFoxp3GFPϪ T cells (effector cells) sorted from 17-day-old SfFoxp3GFP mice expressing respective allelic markers into 7-day-old recipient Ly5.1Ϫ/ϪSfFoxp3GFP mice. When recipi- ent mice were analyzed 10 days after cell transfer, sufficient num- bers of donor cells were only found in the abdominal cavity (site of injection), indicating that donor cells did not undergo expansion in the lymphoreplete environment of recipient mice, consistent with an earlier report (2). In our mouse colony, 10-day-old scurfy

FIGURE 3. Foxp3-deficient Treg cells do not have proliferative advan- mice showed first signs of lymphoproliferative disease and died at tage over effector CD4ϩ T cells and express lower levels of activation ϳ3 wk of age; therefore, a 10-day period of adoptive transfer is markers. A and B, Flow cytometry analysis of lymph node (LN) cells (left appropriate to assess the loss of Foxp3 expression in transferred GFP GFP column) of 3-wk-old Foxp3 and SfFoxp3 mice. Expression of acti- SfFoxp3GFPϩ cells. The great majority (80%) of transferred vation markers CD44 and CD62L on gated conventional (second column) GFPϩ ϩ SfFoxp3 T cells retained Foxp3 transcription; however, their and Treg CD4 cells (third column) is shown. Analysis gates are shown as ϩ proportion in the transferred population decreased (Fig. 5E). Thus, rectangles. C, Percentage of Treg cells in a population of CD4 T cells in Foxp3GFP (B6) and SfFoxp3GFP (Sf) mice are similar. D, The total number this and adoptive transfer experiments described above demon- GFPϩ of CD4ϩ lymph node cells is increased in SfFoxp3GFP (Sf) mice relative to strate that SfFoxp3 cells are much less capable of expansion ϩ Foxp3GFP (B6) mice. than effector CD4 T cells from scurfy mice (see below). Continuous 3736 Foxp3-DEFICIENT Treg CELLS Downloaded from http://www.jimmunol.org/

FIGURE 4. Flow cytometry analysis of lymphocyte population infiltrat- ing peripheral organs in SfFoxp3GFP mice. Panels show expression of the Foxp3GFP reporter in CD4ϩ T cells isolated from liver, lungs, kidney, intestine, and heart. Bar graph shows average percentage of SfFoxp3GFPϩ cells in the respective organs (Li, liver; Lu, lungs; K, kidney; I, intestine, and H, heart). FIGURE 5. Analysis of the proliferative potential and the stability of ϩ the phenotype of Treg and conventional CD4 T cells from healthy and by guest on September 26, 2021 ϩ Ϫ ϩ GFP scurfy mice. A, The fraction of proliferating CD4 GFP and GFP cells transcription of the Foxp3 reporter in vivo and its up-regulation GFP GFP by TGF-␤ treatment in vitro imply that the regulation of Foxp3 was estimated in Foxp3 and SfFoxp3 mice by staining sorted cells with Ki-67-specific Ab (continuous line) or isotype-matched control (bro- expression in Foxp3-deficient and Foxp3-sufficient T cells is Ϫ ϩ reg ken line). B, SfFoxp3GFP CD4 T cells have proliferative advantage over similar and complements earlier reports that inflammatory condi- SfFoxp3GFPϩ cells. Flow cytometry analysis of lymph node cells of TCR ϩ tions preserve Foxp3 expression in most natural or adoptive Treg ␣-chain knockout mice adoptively transferred with total CD4 T cells (5 ϫ cells (data not shown) (2, 33). The origin of a small fraction of 106/mouse) from Foxp3GFPϩ (left panel) and SfFoxp3GFPϩ (right panel) ϩ SfFoxp3GFP T cells that lost Foxp3 expression is not certain. mice. Mice were analyzed 5 wk after transfer. Bar graph shows percentage These cells could represent a subset of genuine, thymus-derived of GFPϩ cells in the CD4ϩ population of Foxp3GFP (bars 1 and 2) and SfFoxp3GFP (bars 3 and 4) cells before transfer (bars 1 and 3) and after Treg cells with less stable phenotype or they might be adoptive Treg (bars 2 and 4) adoptive transfer into lymphopenic mice. C, The fraction of cells that “contaminate” the population of Treg cells. Adoptive Treg proliferating, adoptively transferred CD4ϩGFPϪ and GFPϩ cells in recip- cells do not have a stable Treg phenotype and they can be gener- ated, even in inflammatory conditions, from Foxp3-deficient con- ient mice was estimated by staining cells with Ki-67-specific Ab (contin- uous line) or isotype-matched control (broken line). The results of one of ventional CD4ϩ T cells (34). In conclusion, adoptive transfer GFPϩ three experiments are shown. D, Expression of the GFP reporter in adop- experiments show that SfFoxp3 T cells that lost reporter ex- GFPϩ ϩ Ϫ tively transferred SfFoxp3 CD4 T cells. Flow cytometry analysis of pression may constitute only a very small proportion of GFP cells lymph nodes of TCR ␣-chain knockout recipient mice adoptively cotrans- GFP in SfFoxp3 mice. ferred with sorted lymph node CD4ϩGFPϩLy5.1ϩ cells (2 ϫ 105/mouse) from SfFoxp3GFP mice and total CD4ϩLy5.1Ϫ cells (106/mouse) from GFPϩ ϩ SfFoxp3 CD4 T cells have decreased antigenic response Foxp3GFP mice. Recipient mice were analyzed 4 wk after transfer. E,Ex- and produce a characteristic pattern of cytokines pression of the GFP reporter in adoptively transferred SfFoxp3GFPϩ and SfFoxp3GFPϪ CD4ϩ T cells in the inflammatory environment. CD4ϩ Lower expression of activation markers and decreased homeostatic ϩ/ϩ GFPϩ ϩ ϩ/Ϫ GFPϪ expansion suggested differential reactivity of SfFoxp3GFPϩ cells Ly5.1 SfFoxp3 and CD4 Ly5.1 SfFoxp3 T cells (effector cells) (5 ϫ 105 cells of each subset) sorted from 17-day-old SfFoxp3GFP and effector CD4ϩ T cells. When sorted populations of GFPϩ GFPϪ mice expressing respective allelic markers were transferred i.p. into 7-day- SfFoxp3 and SfFoxp3 cells were stimulated in vitro, pro- old recipient Ly5.1Ϫ/Ϫ SfFoxp3GFP mice. After 10 days, recipient mice GFPϩ liferation of SfFoxp3 cells was very small compared with were sacrificed and cells from the abdominal cavity were analyzed. Flow ϩ effector CD4 T cells and was similar to the proliferation of func- cytometry analysis of gated CD4ϩ of recipient (Ly5.1Ϫ/Ϫ) and donor cells ϩ ϩ/Ϫ ϩ/ϩ tional Foxp3 Treg cells (Fig. 6A). Provision of exogenous IL-2 (Ly5.1 cells, continuous line circle; Ly5.1 cells, gate marked by GFPϩ restored proliferation of both SfFoxp3 and functional Treg dotted line) is shown (left panel). Expression of the GFP reporter is shown ϩ Ϫ Ϫ ϩ ϩ ϩ cells. Thus, SfFoxp3GFPϩ cells closely resemble functional T in Ly5.1 / SfFoxp3GFP (upper right panel) and Ly5.1 / SfFoxp3GFP reg ϩ cells that are anergic in vitro but proliferate and expand in vivo CD4 T cells (lower right panel). Numbers indicate percentages of gated (35). In vitro-activated SfFoxp3GFPϩ and SfFoxp3GFPϪ cells cells. The data show representative data of three recipient mice analyzed. The Journal of Immunology 3737 Downloaded from http://www.jimmunol.org/

FIGURE 6. Functional assays of CD4ϩSfFoxp3GFPϩ T cells. A, SfFoxp3GFPϩ T cells are anergic in vitro. Proliferation of T cell subsets sorted from by guest on September 26, 2021 lymph nodes of Foxp3GFP and SfFoxp3GFP mice is shown. Total CD4ϩ (open bars), CD4ϩFoxp3GFPϪ (dark gray bars), and CD4ϩFoxp3GFPϩ (light gray bars) T cells were stimulated with anti-CD3␧ and anti-CD28 Abs. Addition of rIL-2 (50 U/ml) restored proliferation of CD4ϩFoxp3GFPϩ cells (dotted bars). Data from two individual SfFoxp3GFP (nos.1 and 2) and two Foxp3GFP (nos. 3 and 4) mice are shown. The bars show the average of duplicate measurements. The data represent one of three experiments. B and C, CD4ϩSfFoxp3GFPϩ T cells produce characteristic cytokine pattern. Cytokine levels were measured in supernatants collected from cultures of an equal number of sorted CD4ϩSfFoxp3GFPϪ (upper image) and CD4ϩSfFoxp3GFPϩ (lower image) T cell subsets. Cells were stimulated in vitro for 30 h with plate-bound anti-CD3/anti-CD28 Abs. Chemilumi- nescence images of wells incubated with culture supernatants are shown in B. The arrangement of reagents specific for a particular cytokine is shown in C. MIP-1␣ was abbreviated to MIP1, GM-CSF to GM, and RANTES to RAN. One experiment of three is presented. D, Cytokine levels produced by in vitro-stimulated CD4ϩ T cell subsets measured by ELISA are shown (cytokine concentration is expressed in pg/ml). N.D., Not detectable. E, Analysis of cytokine transcripts present in sorted CD4ϩSfFoxp3GFPϩ and effector SfFoxp3GFPϪ cell subsets of scurfy mice without stimulation in vitro. Cytokine mRNA levels were quantitated by real-time PCR. Relative levels of cytokine mRNA expressed by SfFoxp3GFPϪ (Ϫ, dotted bars) and SfFoxp3GFPϩ (ϩ, dark bars) cells for two mice, A and B are presented in a bar graph. F, Flow cytometry analysis of cytokine production by sorted CD4ϩ T cells from scurfy mice. Cells from healthy Foxp3GFP mice were used as control. Before staining, sorted cells were restimulated in vitro with plate-bound anti-CD3/anti-CD28 Abs for4hinthepresence of monensin.

produced markedly different cytokine profiles (Fig. 6, B–F). Con- sion of Treg cells expressing defective Foxp3 (17). The earlier GFPϩ sistent with proliferation assays, SfFoxp3 cells did not pro- reported level of cytokines produced by Foxp3-deficient Treg duce IL-2. These cells produce low levels of inflammatory cy- cells was much higher than the level of cytokines produced by tokines with the exception of IL-4 and immunoregulatory SfFoxp3GFPϩ cells and led to the conclusion that Foxp3-defi- ␤ GFPϩ cytokines IL-10 and TGF- . SfFoxp3 cells did not differ- cient Treg cells efficiently produced Th1, Th2, and Th17 cyto- entiate into a notable number of Th1 or Th17 helper T cells and kines including IL-2 (16). However, the cells used in the re- did not secrete IL-6 or IL-12, suggesting that they do not indi- ported experiment were treated for a prolonged time with PMA rectly support differentiation of Th17 or Th1 cells, respectively. and ionomycin, whereas we stimulated cells for a short time In contrast, effector CD4ϩ cells produced much higher levels of only with plate-bound Abs. To determine what cytokines are inflammatory cytokines IFN-␥ and MIP-1␣ and a very high produced by T cells isolated directly from scurfy mice, RNA level of IL-4. In addition, SfFoxp3GFPϪ cells produced a expression of IL-2, IL-10, TGF-␤, IFN-␥, IL-4, and IL-17 was broader spectrum of cytokines that included IL-3, IL-5, IL-6, analyzed (Fig. 6E). We detected very low levels of IL-2 tran- IL-10, GM-CSF, and IL-2. The ability to produce a high level scripts in SfFoxp3GFPϩ cells and much lower levels of IFN-␥ of IL-2 is the likely reason for efficient expansion of than in effector CD4ϩ T cells, consistent with intracellular SfFoxp3GFPϪ cells activated by autoantigens in scurfy mice and staining showing that only a small fraction of SfFoxp3GFPϩ explains why these cells are able to support homeostatic expan- cells produces IFN-␥. 3738 Foxp3-DEFICIENT Treg CELLS

FIGURE 7. SfFoxp3GFPϩ T cells do not express el- evated frequency of self-reactive TCRs. A, Analysis of the Foxp3GFP expression in CD4ϩCD25Ϫ and CD4ϩ CD25ϩ T cells shows that both cell subsets include SfFoxp3GFPϩ T cells. Panels in the left column display CD4 and CD25 expression on lymph node cells. Rect- angles show gates used to define CD4ϩCD25Ϫ and CD4ϩCD25ϩ T cell subsets. Foxp3GFP expression on gated CD4ϩCD25Ϫ (middle columns) and CD4ϩ CD25ϩ (right columns) T cells is presented as histo- grams. B, T cell hybridomas produced from SfFoxp3GFPϪ and SfFoxp3GFPϩ CD4ϩ T cells express similar frequency of self-reactive TCRs. Downloaded from SfFoxp3GFPϩ T cells are not enriched in autoreactive CD4ϩ pression of functional Foxp3. To gain further insight into the con- T cells sequences of Foxp3 deficiency for the transcriptional signature of ϩ GFP Treg cells, we have compared the gene expression profiles of reg- Although SfFoxp3 T cells do not exhibit effector functions of ϩ activated conventional CD4ϩ T cells, the low response to antigenic ulatory and conventional CD4 T cells isolated from normal and stimulation in vitro and impaired clonal expansion in vivo may scurfy mice. All gene expression data were obtained from highly conceal their self-reactive potential (5, 36). To investigate self- purified, flow cytometry-sorted cells. Analysis of conventional http://www.jimmunol.org/ GFPϪ GFPϩ reactive T cells, we focused on CD4ϩ T cells that up-regulate Foxp3 and regulatory Foxp3 cells from normal mice activation markers, especially CD25. Up-regulation of CD25 on showed that up-regulation of Gpr83, Folr4, Tnfrsf18 (GITR), ϩ ␣ some SfFoxp3GFP cells may reflect their dependence on IL-2 or, Ctla4, Foxp3, Dusp4, IL-2R , Socs2, and Nrp1 genes are consid- alternatively, denotes their activation status, in particular predis- ered to be a hallmark of the Treg cell transcriptional signature, position to be activated by self-Ags (37). We further investigated consistent with previous reports (Fig. 8A) (16, 17, 30, 38, 39). the relationship between cells expressing CD25 and the GFP re- Some Treg-specific genes (e.g., Gpr83, Folr4, and, Foxp3) were Ϫ GFP up-regulated in Treg cells and down-regulated in conventional porter in SfFoxp3 mice. Our analysis shows that both CD25 ϩ and CD25ϩ T cell subsets have a substantial contribution of CD4 T cells from scurfy and C57BL/6 mice. This suggests that SfFoxp3GFPϩ cells (Fig. 7A). The CD25ϩ T cell subset contained, their expression pattern correlates with cell type, regardless of the by guest on September 26, 2021 GFPϩ Ϫ ability of Treg cells to produce functional Foxp3 protein and re- on average, 2.5–3 times more SfFoxp3 cells than the CD25 ϩ subset. The previously reported greater overlap of the TCR reper- gardless of the activation status of conventional CD4 T cells. It ϩ ϩ toire expressed by presumably autoreactive CD4 CD25 T cells is then possible that high expression of other Treg-specific genes (e.g., CTLA4, Socs2) in the Foxp3-deficient SfFoxp3GFPϩ cell sub- from Foxp3 knockout mice and Treg cells from healthy mice is most likely the consequence of a higher proportion of set is not necessarily a result of cell activation concomitant with GFPϩ ϩ Ϫ SfFoxp3 cells in the population of CD25 than CD25 cells the reversal of the Treg transcriptional program but rather dem- (5). These results also demonstrate that TCRs isolated from the onstrates their persistent Treg phenotype (40). This interpreta- CD25ϩ T cell subset may have originated from SfFoxp3GFPϩ cells tion is consistent with a recent analysis demonstrating that ex- ϩ ϩ instead of autoreactive, expanded, effector CD4 CD25 T cells. pression of some Treg-specific genes overlaps with the TCR To estimate the relative frequencies of self-reactive cells in the response but represents only a subset of the full T cell activation populations of SfFoxp3GFPϩ and SfFoxp3GFPϪ cells, a set of T cell response (41). hybridomas prepared from the respective cell subsets was ana- Remarkably, transcription of Foxp3 was retained in GFPϩ lyzed. This experimental approach allows for analysis of TCR SfFoxp3 cells consistent with RT-PCR analysis. Thus, ex- specificity regardless of the cellular context of a T cell. Sorted pression of the Foxp3 mRNA is regulated by genes located at the GFPϩ GFPϪ SfFoxp3 and SfFoxp3 cells were directly fused, and the higher level of the transcriptional hierarchy of Treg cells and does resulting hybridomas were stimulated with syngenic splenocytes. not require the presence of functional Foxp3 protein. Conse- The frequency of self-reactive TCRs was moderately higher quently, the cellular and molecular features common for ϩ GFPϪ GFP among hybridomas prepared from SfFoxp3 cells (Fig. 7B). SfFoxp3 and Treg cells and described in this report are Foxp3 independent. Functional Foxp3 protein is, however, required to This strongly suggests that although Foxp3-deficient Treg cells may express self-reactive specificities, they are not a major reser- regulate Foxp3-dependent genes. This is illustrated by the expres- voir of autoreactive T cells that might become deleterious upon sion profile of Pde3b (cyclic nucleotide phosphodiesterase 3b) ϩ appropriate stimulation or in patients with dysregulated Foxp3 (Fig. 8A). SfFoxp3GFP cells failed to down-regulate Pde3b ex- function. Hybridoma analysis complements our recent finding that pression as reported earlier (16). the vast majority of regulatory CD4ϩ T cells in healthy mice ex- To reveal the impact of Foxp3 deficiency on the global gene press T cell receptors specific for non-self-ligands (8). signature of Treg cells, we conducted a two-factor ANOVA of gene expression of Foxp3GFPϪ and Foxp3GFPϩ populations isolated The impact of Foxp3 expression on the gene expression profiles from normal mice and SfFoxp3GFPϩ and SfFoxp3GFPϪ populations of Treg cells isolated from scurfy mice. The two factors that influenced the gene GFPϩ The data discussed so far strongly suggest that SfFoxp3 cells expression tested in this analysis were the type of cells (Treg vs ϩ retained important characteristics of Treg cells despite losing ex- conventional CD4 T cell) and the ability to express functional The Journal of Immunology 3739

the level of gene expression in the examined cell type (conven-

tional or Treg cell) is not independent of the other factor, e.g., the ability to express Foxp3. Thus, the genes differentially expressed

in Treg cells from scurfy or healthy mice and showing an interac- tion are the target genes regulated by Foxp3. The result of this analysis is shown in Fig. 8B. Most of the genes presented in the diagram were previously identified as differentially expressed in ϩ Treg or effector CD4 T cells (16, 17, 30, 38, 40–42). Heat maps of these genes are shown indicating where they fall in the Venn diagram. Of the 1039 differentially expressed genes, 183 genes showed a significant interaction, suggesting that they are regulated by Foxp3. This number constitutes 57.7% of all (317) genes differ- ϩ entially expressed in Treg and conventional CD4 T cells. Of the remaining 722 differentially expressed genes, 134 genes were dif- ϩ ferentially expressed in Treg vs conventional CD4 cells regardless of whether the cells were isolated from scurfy or healthy mice, and 759 genes were differentially expressed in cells isolated from

scurfy or healthy mice, regardless of whether they were isolated Downloaded from

from Treg or conventional cells. The set of 134 genes not showing an interaction and differentially expressed in Treg vs conventional ϩ CD4 T cells may represent Treg signature genes independent of Foxp3. The expression pattern of a subset of these genes (97)

remains constant regardless of whether Treg cells from scurfy or

healthy mice are analyzed. Another subset of Treg-specific genes http://www.jimmunol.org/ (37) is differentially expressed in cells isolated from scurfy or

healthy mice. These genes represent Treg-specific genes up- or down-regulated by T cell activation or in response to cytokines,

but whose pattern of expression is the same in Treg and conven- tional CD4ϩ T cells (e.g., they are up- or down-regulated in both subsets). Of 872 genes differentially expressed by CD4ϩ T cells isolated from scurfy or healthy mice, expression of 104 genes was ϩ different in Treg and conventional CD4 T cells. Of these 104 genes, 67 genes showed an interaction indicating dependence on by guest on September 26, 2021 Foxp3. Finally, 722 genes differentially expressed between cells isolated from scurfy and healthy mice did not show dependence on ϩ the cell type (Treg vs conventional CD4 cells) in either scurfy or healthy mice, consistent with the large differences in the number of activated cells. In summary, our analysis demonstrates that a sub-

stantial fraction of the Treg gene signature is controlled by Foxp3; however, the number of these genes is small relative to the number identified using Chip-Chip (18, 19). The origin of CD4ϩ T cells from autoimmune or healthy mice may have a greater quantitative impact on the gene expression profile than the cell type. The list of all genes presented in the Venn diagram is available in supple- mental material.5

Discussion FIGURE 8. Transcriptional profile of Foxp3-deficient and sufficient Treg GFPϩ Recent reports and our own data dissociate the role of Foxp3 in cells. A, Gene expression of Treg-specific genes in Foxp3 and Ϫ ϩ Ϫ Foxp3GFP cells from healthy mice and SfFoxp3GFP and SfFoxp3GFP Treg suppressor function from its role in Treg lineage commitment cells from scurfy mice. B, Venn diagram and expression profiles of genes and provoke new and important questions (16, 17). What is the differentially expressed in Treg vs conventional T cells (blue circle) and scope of Treg cell functions controlled by Foxp3 and what are the ϩ CD4 cells from Foxp3-deficient and sufficient mice (green circle). Red characteristics of Treg cells that lose or down-regulate Foxp3? circle includes genes showing interaction effect, suggesting that they are Since Foxp3 is not critical for the fitness of Treg cells, it is con- Foxp3 dependent. Examples of T -specific genes are shown. Plots inside reg ceivable that some Treg cells may lose or down-regulate Foxp3 the Venn diagram show examples of possible gene expression profiles in expression in healthy individuals. The GFP expression in each section of the diagram. Sf denotes scurfy mice and H denotes healthy Foxp3GFPϩ cells differs 100-fold in healthy mice and correlates B6 mice. with the level of Foxp3 protein expression, suggesting that the Treg population is heterogeneous (Kuczma, M., I. Pawlikowska, M. Ko- pij, R. Podolsky, G. A. Rempala, and P. Kraj, manuscript in prep- Foxp3 protein (scurfy vs C57BL/6). Factorial ANOVA allows us aration). Recent evidence that Foxp3 acts in a dose-dependent, to assess not only the effects of both factors independently of each other but also makes it possible to determine the interaction be- tween factors. In our case, interaction between factors means that 5 The online version of this article contains supplemental material. 3740 Foxp3-DEFICIENT Treg CELLS instead of a binary manner yields further support to the hypothesis conclusion that Foxp3 plays a limited role in the conversion pro- that Treg cells may exist in various shades depending on the level cess (34). In a recent report, the loss of functional Foxp3 expres- of Foxp3 expression (43). Signaling through OX40 may be one of sion led to the reversal of the transcriptional program of Treg cells, ϩ the mechanisms regulating the Foxp3 level and Treg suppressor IL-2 production, and acquisition of the properties of effector CD4 function (44). In this study, we show that Foxp3-deficient Treg T cells (40). Foxp3-deficient Treg cells in male mice lacking func- ϩ cells do not revert to effector CD4 T cells but constitute a distinct tional Treg cells expanded in the periphery and produced IL-2 and subset retaining important cellular characteristics of regulatory Th1, Th2, and Th17 cytokines. Surprisingly, the corresponding cells. cell subset from female mice having a functional population of GFPϩ Peripheral SfFoxp3 cells had a cell surface phenotype dis- Treg cells retained characteristics of Treg cells and remained aner- tinct from conventional T cells and retained features of Treg cells gic to TCR stimulation in vitro, was dependent on IL-2 for pro- despite losing suppressor function. These cells remained depen- liferation, and did not produce inflammatory cytokines (16). Such dent on exogenous IL-2 for proliferation and were anergic in vitro. differences between males and females are difficult to reconcile GFPϩ SfFoxp3 cells produced only small amounts of cytokines with cell autonomous regulation of Treg cell lineage and suggest compared with conventional T cells with the exception of IL-4. that cell-extrinsic factors may convert Treg cells into Teff cells. This corresponds well with molecular findings that the IL-4 Differences in the level of produced cytokines may be due to dif- gene is directly suppressed by Foxp3 (38). The properties of ferent genetic backgrounds of mice used in the previous studies, GFPϩ SfFoxp3 cells in our system closely resemble the properties of distinct stimulation conditions, or difficulties of separating Treg Treg cells expressing a low level of functional Foxp3 that remained cells expressing a low level of GFP from Teff cells. Alternatively, quiescent, produced a Th2-skewed cytokine pattern, and revealed modifications of the endogenous Foxp3 gene to introduce a GFP Downloaded from lower homeostatic expansion than Treg cells expressing a normal reporter or deletions of Foxp3 gene fragments may modify the level of Foxp3 (43). Human Treg cells that down-regulate Foxp3 function of the Foxp3-GFP fusion protein and/or affect detection of expression also tend to produce Th2-type cytokines and convert the Foxp3 transcript. into a Th2 cell type (45). The properties of SfFoxp3GFPϩ cells revealed by cellular and SfFoxp3GFPϩ cells do not differentiate in vitro into a notable immunological analyses correspond well with the analysis of

number of Th1 or Th17 helper T cells and do not secrete high global gene expression using GeneChip technology. The expres- http://www.jimmunol.org/ amounts of IL-6 or IL-12, suggesting that they do not indirectly sion pattern of many signature Treg genes (including Foxp3 itself) support differentiation of Th17 or Th1 cells. Some cytokines, such reported in multiple earlier studies was similar in Foxp3GFPϩ and as GM-CSF, that are important for pathology in scurfy mice are in SfFoxp3GFPϩ cells, suggesting that the imprint of Foxp3 on the produced solely by conventional T cells (16, 46). Analysis of T transcriptional landscape of Treg cells is likely smaller than re- GFPϩ GFPϪ cell hybridomas derived from SfFoxp3 or SfFoxp3 ported earlier (2, 30). To assess the influence of Foxp3 on Treg cell CD4ϩ T cells showed a similar frequency of self-reactive T cells, transcription, we not only relied on comparison of gene expression consistent with our recent report that Treg cells in healthy mice do levels in the relevant cell subsets but determined to what extent not preferentially express self-reactive TCRs (8). Our findings are differential gene expression could be due to the expression of func- by guest on September 26, 2021 consistent with the analysis of Treg expressing a nonfunctional tional Foxp3 protein. This analysis shows that of 317 genes dif- Ϫ Foxp3 mutant. Foxp3-deficient Treg cells produced little IL-2, ferentially expressed between conventional (GFP ) and Treg could not survive when adoptively transferred into recipient mice, (GFPϩ) cells, expression of 183 genes (57.7%) could be regulated and required Foxp3GFPϪ cells to promote autoimmunity (17). by Foxp3. Although a significant fraction of the genes constituting GFPϩ Since SfFoxp3 cells do not produce IL-2, their expansion is the Treg cell transcriptional profile depends on Foxp3, analysis of most likely controlled by the level of IL-2 produced by self-reac- the expression profile of other genes suggests that they are Foxp3 ϩ tive conventional CD4 T cells (47). In conclusion, the features of independent. The existence of a Treg-specific Foxp3-independent SfFoxp3GFPϩ cells do not predispose them to become the dominant set of genes was demonstrated in a recent report (41). The prop- population of self-reactive T cells mediating the fulminant auto- erties of SfFoxp3GFPϩ cells associated with their commitment to a immune disease in scurfy mice. However, the propensity of regulatory lineage are most likely controlled by the set of 97 genes GFPϩ SfFoxp3 cells to produce IL-4 may affect the course of auto- differentially expressed in conventional and Treg cells, not affected immune disease in scurfy mice by augmenting the Th2-type auto- by T cell activation, and independent of Foxp3. In conclusion, immune response and inhibiting generation of Th1 and Th17 cells. gene expression profiling supports our findings that Foxp3-defi- IL-4 was shown to suppress IL-6- and TGF-␤-induced generation cient SfFoxp3GFPϩ cells possess a unique phenotype and do not of Th17 what could explain the absence of these cells in in vitro- revert to conventional effector CD4ϩ T cells. ϩ stimulated CD4 T cells isolated from scurfy mice (48, 49). The The loss of functional Foxp3 protein by Treg cells has dramatic ability to skew the immune response toward Th2 was demon- consequences for the immune system; however, at the level of an strated for Treg cells expressing a low level of functional Foxp3 individual cell, the transition from Foxp3 expression to Foxp3 de- that also have the propensity to produce IL-4 (43). Such modula- ficiency is not associated with a dramatic change in the biology of tion of the autoimmune response may save scurfy mice from the a T cell committed to a regulatory lineage manifested by aug- most destructive tissue damage mediated by Th17 cells (50). mented response to Ag, loss of IL-2 dependence, and production of GFPϩ SfFoxp3 -like cells might be also relevant in chronic autoim- multiple cytokines including IL-2. Rather, Treg cells that lose mune diseases like asthma or allergic diseases where Treg cell de- Foxp3 function become unable to suppress immune responses but ficiency is associated with activation of Th2 effector cells (51). retain production of regulatory and Th2-type cytokines. This has Our data suggest a more limited role of Foxp3 in the lineage important implications for our understanding of autoimmune dis- commitment and differentiation of natural Treg cells than suggested eases and immunotherapeutic approaches. Treg cells that down- earlier (30). Similarly, the scope of cellular functions controlled by regulate or entirely lose Foxp3 expression do not revert to effector ϩ the Foxp3 in adoptive Treg cells seems to be limited to their sup- CD4 T cells that, due to the high frequency of self-reactive T cell pressive function. Adoptive Treg cells produced from conventional receptors, initiate and subsequently dominate autoimmune disease T cells from Foxp3-sufficient and Foxp3-deficient mice revealed but rather constitute a cell subset that modulates effector functions that the gene expression patterns were very similar, leading to the of self-reactive conventional CD4ϩ T cells. The Journal of Immunology 3741

25. Chmielowski, B., P. Muranski, and L. Ignatowicz. 1999. In the normal repertoire Acknowledgments ϩ of CD4 T cells, a single class II MHC/peptide complex positively selects TCRs We thank Dr. N. Copeland from the National Cancer Institute-Frederick for with various antigen specificities. J. Immunol. 162: 95–105. the recombineering reagents. We thank Dr. T. Saunders and M. van Keuren 26. Bolstad, B. M., R. A. Irizarry, M. Astrand, and T. P. Speed. 2003. A comparison from the University of Michigan Transgenic Animal Core for producing of normalization methods for high density oligonucleotide array data based on Foxp3GFP-transgenic mice and Farlyn Hudson for excellent technical variance and bias. Bioinformatics 19: 185–193. 27. Smyth, G. K. 2004. Linear models and empirical bayes methods for assessing assistance. differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. 3: article 3. 28. Brunkow, M. E., E. W. Jeffery, K. A. Hjerrild, B. Paeper, L. B. Clark, Disclosures S. A. Yasayko, J. E. Wilkinson, D. Galas, S. F. Ziegler, and F. Ramsdell. 2001. The authors have no financial conflict of interest. Disruption of a new forkhead/winged-helix protein, scurfin, results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat. Genet. 27: 68–73. 29. Wan, Y. Y., and R. A. Flavell. 2005. Identifying Foxp3-expressing suppressor T References cells with a bicistronic reporter. Proc. Natl. Acad. Sci. USA 102: 5126–5131. 1. Hori, S., T. Nomura, and S. Sakaguchi. 2003. Control of regulatory T cell de- 30. Fontenot, J. D., J. P. Rasmussen, L. M. Williams, J. L. Dooley, A. G. Farr, and velopment by the transcription factor Foxp3. Science 299: 1057–1061. A. Y. Rudensky. 2005. Regulatory T cell lineage specification by the forkhead 2. Fontenot, J. D., M. A. Gavin, and A. Y. Rudensky. 2003. Foxp3 programs the transcription factor Foxp3. Immunity 22: 329–341. ϩ ϩ development and function of CD4 CD25 regulatory T cells. Nat. Immunol. 4: 31. Gerdes, J., H. Lemke, H. Baisch, H. H. Wacker, U. Schwab, and H. Stein. 1984. 330–336. Cell cycle analysis of a cell proliferation-associated human nuclear antigen de- 3. Khattri, R., T. Cox, S. A. Yasayko, and F. Ramsdell. 2003. An essential role for fined by the monoclonal antibody Ki-67. J. Immunol. 133: 1710–1715. ϩ ϩ Scurfin in CD4 CD25 T regulatory cells. Nat. Immunol. 4: 337–342. 32. Hori, S., M. Haury, J. J. Lafaille, J. Demengeot, and A. Coutinho. 2002. Periph- 4. Jordan, M. S., A. Boesteanu, A. J. Reed, A. L. Petrone, A. E. Holenbeck, eral expansion of thymus-derived regulatory cells in anti-myelin basic protein T ϩ ϩ M. A. Lerman, A. Naji, and A. J. Caton. 2001. Thymic selection of CD4 CD25 cell receptor transgenic mice. Eur. J. Immunol. 32: 3729–3735. regulatory T cells induced by an agonist self-peptide. Nat. Immunol. 2: 301–306. 33. Huter, E. N., G. A. Punkosdy, D. D. Glass, L. I. Cheng, J. M. Ward, and 5. Hsieh, C. S., Y. Zheng, Y. Liang, J. D. Fontenot, and A. Y. Rudensky. 2006. An E. M. Shevach. 2008. TGF-␤-induced Foxp3ϩ regulatory T cells rescue scurfy Downloaded from intersection between the self-reactive regulatory and nonregulatory T cell recep- mice. Eur. J. Immunol. 38: 1814–1821. tor repertoires. Nat. Immunol. 7: 401–410. 34. Haribhai, D., W. Lin, B. Edwards, J. Ziegelbauer, N. H. Salzman, M. R. Carlson, 6. van Santen, H. M., C. Benoist, and D. Mathis. 2004. Number of Treg cells that S. H. Li, P. M. Simpson, T. A. Chatila, and C. B. Williams. 2009. A central role differentiate does not increase upon encounter of agonist ligand on thymic epi- for induced regulatory T cells in tolerance induction in experimental colitis. J. Im- thelial cells. J. Exp. Med. 200: 1221–1230. munol. 182: 3461–3468. 7. Wong, J., R. Obst, M. Correia-Neves, G. Losyev, D. Mathis, and C. Benoist. 35. Klein, L., K. Khazaie, and H. von Boehmer. 2003. In vivo dynamics of antigen- 2007. Adaptation of TCR repertoires to self-peptides in regulatory and nonregu- specific regulatory T cells not predicted from behavior in vitro. Proc. Natl. Acad. ϩ

latory CD4 T cells. J. Immunol. 178: 7032–7041. Sci. USA 100: 8886–8891. http://www.jimmunol.org/ 8. Pacholczyk, R., J. Kern, N. Singh, M. Iwashima, P. Kraj, and L. Ignatowicz. 36. Hsieh, C. S., Y. Liang, A. J. Tyznik, S. G. Self, D. Liggitt, and A. Y. Rudensky. ϩ 2007. Nonself-antigens are the cognate specificities of Foxp3 regulatory T cells. 2004. Recognition of the peripheral self by naturally arising CD25ϩCD4ϩ T cell Immunity 27: 493–504. receptors. Immunity. 21: 267–277. 9. Paust, S., and H. Cantor. 2005. Regulatory T cells and autoimmune disease. 37. Fehervari, Z., T. Yamaguchi, and S. Sakaguchi. 2006. The dichotomous role of Immunol. Rev. 204: 195–207. IL-2: tolerance versus immunity. Trends Immunol. 27: 109–111. 10. Huan, J., N. Culbertson, L. Spencer, R. Bartholomew, G. G. Burrows, 38. Sugimoto, N., T. Oida, K. Hirota, K. Nakamura, T. Nomura, T. Uchiyama, and Y. K. Chou, D. Bourdette, S. F. Ziegler, H. Offner, and A. A. Vandenbark. 2005. S. Sakaguchi. 2006. Foxp3-dependent and -independent molecules specific for Decreased Foxp3 levels in multiple sclerosis patients. J. Neurosci. Res. 81: CD25ϩCD4ϩ natural regulatory T cells revealed by DNA microarray analysis. 45–52. Int. Immunol. 18: 1197–1209. 11. Balandina, A., S. Lecart, P. Dartevelle, A. Saoudi, and S. Berrih-Aknin. 2005. 39. Yamaguchi, T., K. Hirota, K. Nagahama, K. Ohkawa, T. Takahashi, T. Nomura, Functional defect of regulatory CD4ϩCD25ϩ T cells in the thymus of patients and S. Sakaguchi. 2007. Control of immune responses by antigen-specific regu-

with autoimmune myasthenia gravis. Blood 105: 735–741. latory T cells expressing the folate receptor. Immunity 27: 145–159. by guest on September 26, 2021 12. Levy-Lahad, E., and R. S. Wildin. 2001. Neonatal diabetes mellitus, enteropathy, 40. Williams, L. M., and A. Y. Rudensky. 2007. Maintenance of the Foxp3-depen- thrombocytopenia, and endocrinopathy: further evidence for an X-linked lethal dent developmental program in mature regulatory T cells requires continued ex- syndrome. J. Pediatr. 138: 577–580. pression of Foxp3. Nat. Immunol. 8: 277–284. 13. Godfrey, V. L., J. E. Wilkinson, and L. B. Russell. 1991. X-linked lymphore- 41. Hill, J. A., M. Feuerer, K. Tash, S. Haxhinasto, J. Perez, R. Melamed, D. Mathis, ticular disease in the scurfy (sf) mutant mouse. Am. J. Pathol. 138: 1379–1387. and C. Benoist. 2007. Foxp3 transcription-factor-dependent and -independent 14. Clark, L. B., M. W. Appleby, M. E. Brunkow, J. E. Wilkinson, S. F. Ziegler, and regulation of the regulatory T cell transcriptional signature. Immunity 27: F. Ramsdell. 1999. Cellular and molecular characterization of the scurfy mouse 786–800. mutant. J. Immunol. 162: 2546–2554. 42. Knoechel, B., J. Lohr, S. Zhu, L. Wong, D. Hu, L. Ausubel, and A. K. Abbas. 15. Blair, P. J., S. J. Bultman, J. C. Haas, B. T. Rouse, J. E. Wilkinson, and 2006. Functional and molecular comparison of anergic and regulatory T lym- V. L. Godfrey. 1994. CD4ϩCD8 ϪT cells are the effector cells in disease patho- phocytes. J. Immunol. 176: 6473–6483. genesis in the scurfy (sf) mouse. J. Immunol. 153: 3764–3774. 43. Wan, Y. Y., and R. A. Flavell. 2007. Regulatory T-cell functions are subverted 16. Gavin, M. A., J. P. Rasmussen, J. D. Fontenot, V. Vasta, V. C. Manganiello, and converted owing to attenuated Foxp3 expression. Nature 445: 766–770. J. A. Beavo, and A. Y. Rudensky. 2007. Foxp3-dependent programme of regu- 44. Vu, M. D., X. Xiao, W. Gao, N. Degauque, M. Chen, A. Kroemer, N. Killeen, latory T-cell differentiation. Nature 445: 771–775. N. Ishii, and X. C. Li. 2007. OX40 costimulation turns off Foxp3ϩ Tregs. Blood 17. Lin, W., D. Haribhai, L. Relland, N. Truong, M. Carlson, C. Williams, and 110: 2501–2510. T. Chatila. 2007. Regulatory T cell development in the absence of functional 45. Veldman, C., A. Pahl, S. Beissert, W. Hansen, J. Buer, D. Dieckmann, Foxp3. Nat. Immunol. 8: 359–368. G. Schuler, and M. Hertl. 2006. Inhibition of the transcription factor Foxp3 con- 18. Marson, A., K. Kretschmer, G. M. Frampton, E. S. Jacobsen, J. K. Polansky, verts desmoglein 3-specific type 1 regulatory T cells into Th2-like cells. J. Im- K. D. Macisaac, S. S. Levine, E. Fraenkel, B. H. von, and R. A. Young. 2007. munol. 176: 3215–3222. Foxp3 occupancy and regulation of key target genes during T-cell stimulation. 46. Kim, J. M., J. P. Rasmussen, and A. Y. Rudensky. 2007. Regulatory T cells Nature 445: 931–935. prevent catastrophic autoimmunity throughout the lifespan of mice. Nat. Immu- 19. Zheng, Y., S. Z. Josefowicz, A. Kas, T. T. Chu, M. A. Gavin, and nol. 8: 191–197. A. Y. Rudensky. 2007. Genome-wide analysis of Foxp3 target genes in devel- 47. Bayer, A. L., A. Yu, and T. R. Malek. 2007. Function of the IL-2R for thymic and oping and mature regulatory T cells. Nature 445: 936–940. peripheral CD4ϩCD25ϩ Foxp3ϩ T regulatory cells. J. Immunol. 178: 20. Campbell, D. J., and S. F. Ziegler. 2007. FOXP3 modifies the phenotypic and 4062–4071. functional properties of regulatory T cells. Nat. Rev. Immunol. 7: 305–310. 48. Harrington, L. E., R. D. Hatton, P. R. Mangan, H. Turner, T. L. Murphy, 21. Lopes, J. E., T. R. Torgerson, L. A. Schubert, S. D. Anover, E. L. Ocheltree, K. M. Murphy, and C. T. Weaver. 2005. Interleukin 17-producing CD4ϩ effector H. D. Ochs, and S. F. Ziegler. 2006. Analysis of Foxp3 reveals multiple domains T cells develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat. required for its function as a transcriptional repressor. J. Immunol. 177: Immunol. 6: 1123–1132. 3133–3142. 49. Park, H., Z. Li, X. O. Yang, S. H. Chang, R. Nurieva, Y. H. Wang, Y. Wang, 22. Gavin, M. A., T. R. Torgerson, E. Houston, P. deRoos, W. Y. Ho, L. Hood, Z. Zhu, Q. Tian, and C. Dong. 2005. A distinct lineage of CD4 T cells A. Stray-Pedersen, E. L. Ocheltree, P. D. Greenberg, H. D. Ochs, and regulates tissue inflammation by producing interleukin 17. Nat. Immunol. 6: A. Y. Rudensky. 2006. Single-cell analysis of normal and FOXP3-mutant human 1133–1141. T cells: FOXP3 expression without regulatory T cell development. Proc. Natl. 50. Stummvoll, G. H., R. J. DiPaolo, E. N. Huter, T. S. Davidson, D. Glass, Acad. Sci. USA 103: 6659–6664. J. M. Ward, and E. M. Shevach. 2008. Th1, Th2, and Th17 effector T cell-induced 23. Wang, J., A. Ioan-Facsinay, d. van, V., T. W. Huizinga, and R. E. Toes. 2007. autoimmune gastritis differs in pathological pattern and in susceptibility to sup- Transient expression of FOXP3 in human activated nonregulatory CD4ϩ T cells. pression by regulatory T cells. J. Immunol. 181: 1908–1916. Eur. J. Immunol. 37: 129–138. 51. Hartl, D., B. Koller, A. T. Mehlhorn, D. Reinhardt, T. Nicolai, D. J. Schendel, 24. Warming, S., N. Costantino, Court, D. L., N. A. Jenkins, and N. G. Copeland. M. Griese, and S. Krauss-Etschmann. 2007. Quantitative and functional impair- 2005. Simple and highly efficient BAC recombineering using galK selection. ment of pulmonary CD4ϩCD25high regulatory T cells in pediatric asthma. Nucleic Acids Res. 33: e36. J. Allergy Clin. Immunol. 119: 1258–1266. Supplemental Material

List of genes presented on the Venn diagram in Figure 8

Interaction – Treg/Teff – Sf/WT Cd83 Melk Afp Brap Ccnd1 Cdr2 Pdlim4 Nrp1 Ggtla1 Socs2 Sag Cst7 Pip5k2a Ctla4 Il1r2 Ccl3 Tnfrsf4 Gbp1 Il2ra Aebp2 Ccl1 Acss2 Pdk1 Atp1b1 Mtg1 Gprc5b Sccpdh Qpct Mki67 Map2k6 Penk1 1110038D17Rik Tpx2 Osbpl3 Myo1e 2310010M24Rik 4930524L23Rik A130092J06Rik Endod1 Ntrk3 D16Bwg1543e Slc16a5 Amigo2 Ampd1 Acvrl1 A730095J18Rik Cnn3 AI790276 Tcrb-V13 Igf1r Prnp Cish Car2 Il2rb Slc12a7 Ebi3 Pdcd1lg2 Arhgap12 Acvrl1 Cnga1 Actn1 Gins1 Stx11 Arhgap29 Xkrx Sema7a Cybrd1

Interaction – Sf/WT Lxn Tspan31 Tmbim1 Itk Unc119 Il12rb1 Atp1b1 Bcl2a1a Gzmb Dnajc3a S100a6 BC003324 Prr13 S100a4 Tyms Hip1r 2810439F02Rik Tgfbr2 Katnb1 Chst11 Nuf2 Zbtb32 Synpo AW112010 Tjp2 Csda Icos Sh3bgrl Kif22 Bace2 Tmem180 Mcm6 Myadm Treml2 Bub1b Casp3 Tjp2 Csda Atp1b1 Zdhhc2 Pebp1 Plscr1 Exosc1 St8sia6 5930435M05Rik Naip2

Interaction – Treg/Teff Mllt11 Cyfip1 Gsto1 Mettl9 Tnfrsf1b Decr1 Myo1c Il2ra Foxp3 Klrg1 Enc1 Tnfrsf18 Niban Pim1 Mmd Lrrc61 Plagl1 App 2610019F03Rik Gramd3 Dusp4 Tcf7 Pde3b Tgfbr3 Hs3st3b1 Lycat Il1rl2 Ikzf1 E430004N04Rik Itgae Lrmp Nrp1 Socs2 Enc1 Capg Coro2a Galm AU019833 Rgs16 AV071699 Dse 1500005K14Rik Mctp2 Tnfrsf9

Interaction Satb1 Rcbtb2 Emp1 Nfe2l2 Syngr2 Srprb Mxd4 Tmem173 Blr1 Hmgn1 Atp1b3 A530021J07Rik Smurf2 Mobkl2b Slc2a3 Igfbp4 Klhdc1 A530021J07Rik Ctnna1 Pdcd1 Dbndd2 Myo10 Syt11 Commd8 Slc2a3 S100a10

Treg/Teff – Sf/WT Adh1 Nusap1 H2afz Glrx Mcl1 Rgs1 Tiam1 Adam19 1500005A01Rik Itgax Mbnl3 Rab6ip1 Eya2 Smyd2 Coro2a Rapgef4 Ptprs ENSMUSG00000068790 Paox Cnksr3 Cerk Lass6 Eomes Cdk6 C920005C14Rik Dbp Rpap1 Casp7 Hpcal1 Hsd11b1 Prg4 F2r Nedd4 Alox5ap Skil Ncoa7 Rtn4rl1

Sf/WT Uhrf1 Stmn1 Kpna2 Rrm1 Mgst1 Lpl Tpi1 Marcksl1 Nup62 Cox6c Carhsp1 Tspan3 Ehd1 Fabp5 Fabp5 Mcm7 Mcm7 Pfkp Ccnb1-rs1 Adam9 Txn1 Rrm2 Fkbp5 Hmgb3 Prdx4 Mcm4 Mcm6 Tk1 Shcbp1 Litaf Litaf Crip1 Snrpd1 Cd82 Rpa2 Ccne1 Ezh2 Cdc45l Bad Glrx Id3 Lig1 Cdc20 Coq7 Cks1b Hmg20a H2afx Zwilch Wee1 Tdg Cdca5 Ctla2a Cdkn2c Adam8 Atp6v1d Bub1b Rcc1 Foxo1 Foxo1 Cdc6 Orc6l Anp32b Ahcy Srpk2 Pim2 Ccl6 Nek2 Smpdl3b Psrc1 Peli1 C1qdc2 Ndc80 Tacc3 Cks2 Cks2 Ifitm2 Apobec3 Fbxo9 Gmnn Rnf19 Srp9 Hells Mybl2 Got2 Serinc3 D17H6S56E-5 Pgk1 Gstt2 Gzma Ccna2 Ccna2 Ncapg2 Ccl9 Rad51ap1 Bhlhb2 Exo1 Prim2 Prim2 Ccl5 Usp18 Galnt10 Cenpk Rad51 Dbf4 Prim1 Dynll2 Tspan32 Mir16 Exosc9 Fgf13 Stambpl1 Dgka Nr2c1 Egln3 Egln3 Cxcl9 Sgol1 Cxcl10 Bak1 Daxx Anxa2 2810417H13Rik 2810417H13Rik Dhfr Rgs14 Dut H2-Oa Sorcs2 Pola1 Il10rb Nudc Ect2 Lgals1 Ms4a6d Ms4a6d Clec4n Ier3 Chi3l3 S100a1 Plk4 Ccnb1 Hdgf Mcm3 Cd160 D2Ertd750e Ccl6 Lilrb4 Nap1l1 Vps26a Anp32e Grap2 Slc15a3 Traip Ppp3cc 4930583H14Rik Art2b Rsad2 Acyp1 Dnase1l3 Jak2 Foxp1 Foxp1 Foxp1 Lif Anxa4 Ugcg Isgf3g Racgap1 Sytl2 Fen1 Kcnk5 Fgl2 Serhl Cdc25b Cenph H2-T24 Cyb5r3 Fignl1 Ephx1 Xpnpep1 Itga6 Mad2l1 Ube2t Ccnf H2-DMa Ccne2 Mt1 Igf2bp3 Igf2bp3 Ahr Slc43a3 Serpinb6b Cxcr6 Aspm Dnmt1 Txnl5 Incenp Lrrc41 Lrrc41 Tfdp1 Rps27l Polg Ccdc16 Osbpl9 1110001A07Rik Lmnb1 Lmnb1 Paics Ddx39 Cdipt Ela1 Rfc3 H1f0 Asf1b Ak3 Prc1 Prc1 Tcf19 Kif22 Ncapd2 Rfwd3 Ppp2r5a Cttn Ncaph C3 Nup37 Thoc3 Hint1 Tmem66 Bub1 Dera Neil3 Rtcd1 Kif18a Spc25 Aurkb Ppih Cdt1 1500011H22Rik Ifitm1 Smox Birc5 Depdc1a Rfc4 Tmem140 Asrgl1 Armc7 Aurka Slc25a4 Brca1 Cdkn1a 1810015C04Rik Runx2 Ercc6l Oas1g Fh1 Hist1h4i Trim34 Zbtb24 Alad Gpsm2 Serpina3g Ccdc5 Ccdc99 Il1rl1 Shmt1 Igh-6 Igh-6 Igh-6 Hmmr Ifng Clec4n Mapre2 Mapre2 Alcam Siva1 Tagln2 Pgam1 3100002L24Rik Dis3 Tipin Fn1 Mapkapk2 Mcm3 Cnn3 Prmt3 Lgals3 Pold3 Scrn3 Itgb1 Trim33 Rnf43 Rbbp8 Cenpf Atp2a2 Fxn Suv39h1 Csf2 Hmmr Stil Igh-6 Igh-6 Pmvk Cdca7 Tpx2 Lman1 Tspan9 Esco2 Larp7 Agpat4 Cdca8 Cdca8 2610528E23Rik D1Bwg0212e Pak4 Tbc1d9b P2ry5 Ccdc128 Prr6 Gins2 Whsc1 Atm Mt2 Tmem107 Cenpp 2610036L11Rik Ncapg 2610524H06Rik 1190007I07Rik Syce2 Trip13 Rhoh Cenpl Ggps1 4930579G24Rik 2010317E24Rik Cpm 1110002L01Rik Plscr1 Ppp3cc Rab19 2310031A07Rik Rnf167 Nmral1 Cdkn3 Farsb Peci Ccdc50 Spc24 Zfp655 LOC677168 Tube1 3000002C10Rik Cenpp Mcm10 Cdk5r1 Cdk5r1 AI467606 AI467606 6430527G18Rik Aqr LOC100039489 Inpp5f Anln Acad11 Dgkd Tloc1 Sesn1 Igf2bp3 Trio Slc31a1 Slc39a10 Snhg3 Jak1 Espl1 Spag5 Spag5 Cttn Myo6 Asns Usp28 Znrf2 Repin1 AI662270 Mettl7a Mettl7a 4632434I11Rik Fas Rasgrp1 Rab3ip Chsy1 Prkd2 BC055324 Cd44 Galnt6 Rnf157 Rrm2 Sp100 Ugp2 Rapgef6 2410066E13Rik Tjp2 Dtl Mga Ckap2 Garnl4 C79407 Etv6 Ppp1r13b Slc25a4 Arhgap19 Hirip3 Cenpe Eme1 Pitpnc1 Klhdc2 Ugcg Sft2d2 Id2 Ibrdc3 Btaf1 Itpkb Pyhin1 Rhoc Marcksl1 Epas1 Bcl2l11 Ttc28 Fcgr2b Cdk2ap1 Kntc1 Atad5 Marcksl1 Tpi1 Rhobtb2 LOC624784 Plcb4 4930547N16Rik Maf LOC625360 Sfxn2 Ckap2l Gdi2 Whsc1 Itga4 Rsad2 AW011738 Samd9l E2f8 2700094K13Rik Card6 1700097N02Rik 9230110J10 Fen1 Zfp260 Ddx58 Prkd2 Ncaph Mcm4 Pif1 Mcm5 Cdca8 Skp2 E2f7 Marcksl1 Tmem154 F2r Hmgb2 Tgfbi Efhd2 1110001A07Rik B430306N03Rik Mcoln3 Nek2 Kif2c Phgdh Xrcc3 E030037K03Rik Nap1l1 Herc5 H2afz Vim Rfc4 Frmd4b Pmf1 Hmgb2 Mcm7 Arhgap11a Eif4h St8sia6 Bub1 Clic4 Cryz Pgk1 Bcl11b Ywhae Phf11 Sesn1 Txnl5 Gpr137b-ps Abhd4 Mcm7 Ran Cdc20 C1qdc2 Incenp Dsg2 Sgol1 F730047E07Rik C330018D20Rik Tmem170 Sfxn2 AI851523 Psmd11 Sfxn2 Bri3 Ccne1 BC033915 Abce1 Hsh2d H2-DMb2 H2-DMb2 Fbxo39 Ccnf Cd55 LOC100043424 Stc2 Arl1 Foxp1 LOC100039220 Pde2a 5730469M10Rik Maf Gapdh Baz2b Stmn1 Tgfbi Vcam1 Plk1 Ccnb1-rs1 Anxa1 Rrm2 Crmp1 Cdc2a 4632428N05Rik Cks1b Cdca5 Ctla2a Slmo2 Srpk2 Pbk Smc2 Wdr12 Pole 3110048E14Rik Plek Atpif1 Ifi203 Moap1 Mxd3 Rad51ap1 Rbpj Ifit3 Prim1 Nudt4 Kif20a Sostdc1 Klf7 Ttc28 Bcl7c Ebi2 C330027C09Rik Mcm3 Chek1 Il4 Epas1 Nqo2 Sgk3 Acyp1 Serpinb6a Hmmr Fmo5 2810433K01Rik Gng10 Kif4 Fkbp2 Rdbp Ell2 Nkg7 Ifit1 Saa3 Cenpa Calm3 Bcat1 Lipa Mpp1 Ccnb2 Osbpl9 Stk17b Psat1 Asns D14Ertd500e Dap Cmtm6 Kif22 Slc39a4 Spcs3 Dnajb4 Aurkb Racgap1 Atp11b Cbr4 Gemin8 Parp8 6720467C03Rik Havcr2 Slfn8 Fcgr2b Lrrk1 Mapre2 Cdca3 Chtf18 Frmd4b Ptpn13 Dna2l Ahnak Cep55 Cenpn Kif11 Ctla2b AA881470 Dgcr6 Ppil5 Cd44 Kit Hmgb2 Hist1h2bp Itgb1 Zdhhc2 Tfrc Rps4y2 Pole3 5730494N06Rik Glipr2 1500009L16Rik Il6st 2700029M09Rik Gins2 2600005O03Rik Rfc5 Tpi1 Ccdc28b Ube2c Rnaseh2b Cpm 1810054D07Rik Slamf7 Plp2 Herc5 Psat1 6430527G18Rik Cbx5 Top2a Sesn1 Ablim1 F630043A04Rik Rxra AW548124 Mettl7a 1110013L07Rik Rnf157 Gnai1 Hsdl1 Slc25a4 Foxp1 Smarcc1 Fcgr2b Lgals1 Pdzk1ip1 Rnf213 Serinc3 Samd9l LOC320895 Csf2rb 1700097N02Rik Dlg7 Tacc3 Scpep1 Cdca2 Kif23 H2afz Pml St8sia6 2700049P18Rik Tgfbi Clspn Slfn5 Vim Il18rap Gdi2 2900073G15Rik Thada Amy2 Zfp78 Brip1 LOC100039220 Ccdc28b H2-DMa 2700007P21Rik Sod1 Slbp Cd55 Tnfsf13b Hist2h3c1 Sh3kbp1 S100a11 Tmem160 Gcnt1 BC033915 Samd9l D15Wsu75e Gapdh

Treg/Teff Emb Emb Gsta4 Afp Ubac1 Tbc1d8 Arl2bp Il2rb Pscd3 Ccdc109b Sema4f Lta Folr4 Birc3 Tcrg Mtap6 Tmepai Ergic1 Gpr83 Dst Tspyl4 Igf2r BC020489 Slamf1 Slamf1 Maged2 Appl2 C230093N12Rik Mdfic Mpeg1 4833442J19Rik Tmem158 Oxct1 Cyp2s1 Nt5e Hod Itih5 Rabgap1l Btbd14a 2810474O19Rik 1500005K14Rik Rnase6 Ndg2 Tmem64 Arl5a Mtmr3 Prkch Tmem64 Prkar1b 4732460K03Rik Lyst Mxd1 Asxl1 Pim1 Trim14 Itgb8 Arl4c Ikzf3 Itih5 Afp Ndg2 D630039A03Rik Mettl9 Ikzf4 Gpr171 Dnahc7b 1190002H23Rik Sema4a Gpr83 D13Ertd608e P2rx7 Atp8b4 A630014C17Rik Tmie Acpp Tmie Klk1 Il7r Ecm1 St3gal6 Gata1 Gadd45b Crtam Prf1 4631427C17Rik ENSMUSG00000068790 Tmem64 Cpd 2810001G20Rik Itgb3 AI449310 Gsto1 1110032F04Rik Acpl2 Ikzf4 Pde7a Btbd11