Functional and Molecular Comparison of Anergic and Regulatory T Lymphocytes Birgit Knoechel, Jens Lohr, Shirley Zhu, Lisa Wong, Donglei Hu, Lara Ausubel and Abul K. Abbas This information is current as of September 29, 2021. J Immunol 2006; 176:6473-6483; ; doi: 10.4049/jimmunol.176.11.6473 http://www.jimmunol.org/content/176/11/6473 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2006/05/18/176.11.6473.DC1 Material References This article cites 54 articles, 28 of which you can access for free at:

http://www.jimmunol.org/content/176/11/6473.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

by guest on September 29, 2021 • 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 © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Functional and Molecular Comparison of Anergic and Regulatory T Lymphocytes1

Birgit Knoechel,2* Jens Lohr,2* Shirley Zhu,† Lisa Wong,† Donglei Hu,† Lara Ausubel,3* and Abul K. Abbas4*

Tolerance in vivo is maintained by multiple mechanisms that function to prevent autoimmunity. An encounter of CD4؉ T cells with a circulating self-Ag leads to partial thymic deletion, the development of CD25؉ regulatory T cells (Tregs), and functional anergy in the surviving CD25؊ population. We have compared anergic and regulatory T cells of the same Ag specificity generated in vivo by the systemic self-Ag. Anergic cells are unresponsive to the self-Ag that induces tolerance, but upon transfer into a new host and immunization, anergic cells can induce a pathologic autoimmune reaction against tissue expressing the same Ag. Tregs, in contrast, are incapable of mediating harmful reactions. To define the basis of this functional difference, we have compared expression profiles of anergic and regulatory T cells. These analyses show that Tregs express a distinct molecular signature, but Downloaded from anergic cells largely lack such a profile. Anergic cells express transcripts that are associated with effector differentiation, e.g., the effector cytokines IL-4 and IFN-␥. Anergic cells do not produce these cytokines in response to self-Ag, because the cells exhibit a proximal signaling block in response to TCR engagement. Thus, anergy reflects an aborted activation pathway that can readily be reversed, resulting in pathologic effector cell responses, whereas Treg development follows a distinct developmental pathway that extinguishes effector functions. The Journal of Immunology, 2006, 176: 6473–6483. http://www.jimmunol.org/

cell tolerance to self-Ags is maintained by three principal To address these questions, we have generated a Tg mouse that mechanisms: anergy (functional unresponsiveness), dele- expresses a secreted form of OVA (sOVA) in the circulation T tion (apoptotic death), and suppression by regulatory T (sOVA Tg mouse) and crossed these animals with TCR Tg mice lymphocytes (1–7). Many of the studies on which this conclusion expressing the DO11.10 (referred to hereafter as DO11) OVA- is based have relied on cloned cell lines or polyclonal stimuli (8– specific TCR. This is the first experimental system in which both 10). Much less has been done to address the mechanisms of T cell anergic and regulatory T cells are induced by a single self-Ag in tolerance induced in normal T cells by self-Ags, especially in vivo. the same Ag-specific population, and thus it is possible to formally 5 by guest on September 29, 2021 Transgenic (Tg) mouse models are valuable for defining the fates compare the two cell populations. In this work we show that an- of T cells that encounter their cognate Ag in different forms. A ergic and regulatory T cells can be distinguished on the basis of large number of studies have shown that deletion, anergy, and function and profiles. Microarray analysis re- ϩ suppression by CD25 regulatory T cells (Tregs) are all demon- vealed that anergic CD4ϩ T cells express very few specific . strable with Ag receptor Tg T cells (11–17). However, few studies However, they express abundant mRNA for cytokine genes such have formally compared the properties of anergic and regulatory T as Ifn-␥ and Il4 that are typically assigned to Th cells with effector cells induced in one T cell population by one Ag. In particular, function, suggesting that anergic cells in this setting may be given the considerable interest in inducing T cell anergy as a ther- “poised” to become true effector cells. We demonstrate that this apeutic strategy, it is important to ask whether this may be more or differentiation also occurs if tolerance is induced de novo when T less effective than generating Tregs. It is also unclear whether the cells newly encounter a tolerogenic Ag in an adoptive transfer biochemical and molecular characteristics of anergic and regula- model (18). Similar to T cells that have encountered self-Ag during tory T cells are distinct or show some overlap. development, the transferred T cells express IFN-␥ mRNA within four days after encounter with the self-Ag. Anergic T cells can *Department of Pathology and †Diabetes Center, University of California San Fran- even be triggered to secrete IFN-␥ upon stimulation in vivo. How- cisco School of Medicine, San Francisco, California 94143 ever, at the same time anergic cells develop a proximal signaling ϩ Received for publication August 9, 2005. Accepted for publication March 15, 2006. defect leading to decreased Ca2 mobilization upon TCR stimu- The costs of publication of this article were defrayed in part by the payment of page lation. Importantly, we show that the state of anergy can be broken charges. This article must therefore be hereby marked advertisement in accordance in that anergic T cells are unresponsive to the systemic self-Ag but with 18 U.S.C. Section 1734 solely to indicate this fact. retain their ability to trigger severe pathologic autoimmunity. In 1 This work was supported by National Institutes of Health Grant PO1 AI35297 and Deutsche Forschungsgemeinschaft Fellowships KN 533/1-1 and LO 808/1-1. contrast, regulatory T cells are incapable of causing harmful au- 2 These authors contributed equally to this work. toimmune reactions, and they express a large number of genes that 3 Current address: City of Hope Medical Center, Duarte, CA 91010. are specific to this lineage. Our studies suggest that anergy induc- tion is a double-edged sword in which a signaling block and 4 Address correspondence and reprint requests to Dr. Abul K. Abbas, Department of Pathology, University of California San Francisco, 505 Parnassus Avenue, M590, San aborted effector responses develop simultaneously and that this Francisco, CA 94143-0511. E-mail address: [email protected] balance may be easily tipped. Regulatory T cells, in contrast, are 5 Abbreviations used in this paper: Tg, transgenic; HPRT, hypoxanthine phosphori- incapable of effector responses, suggesting that they develop by a bosyltransferase; mOVA, membrane-bound OVA; sOVA, soluble form of OVA; RIP, rat insulin promoter; RMA, robust multiarray average; Treg, CD25ϩ regulatory T distinct pathway. These data have implications for a variety of cell. approaches attempting to induce tolerance in vivo.

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00 6474 ANERGIC AND REGULATORY T LYMPHOCYTES

Materials and Methods from DO11 mice were used as responders and stimulated on 25,000 APCs with 1 ␮g/ml OVA. The input of responder cells remained constant. KJ1- Mice ϩ ϩ ϩ ϩ ϩ Ϫ 26 CD4 CD25 and KJ1-26 CD4 CD25 T cells from double-Tg mice All experimental mice were used at 6–12 wk of age. All mice were age and were used as suppressors and, starting from equal numbers, were titrated in sex matched Ϯ2 wk. BALB/c mice were purchased from Charles River 1/4 dilutions. [3H]Thymidine (1 ␮Ci/well) was added during the last 16 h Laboratory. Tg mice expressing the DO11 TCR specific for the chicken of culture, and incorporation was measured by scintillation counting after

OVA peptide (OVA323–339) in the context of the MHC class II molecule 72 h of culture. I-Ad were obtained from K. Murphy (Washington University, St. Louis, MO). RNA purification and amplification sOVA Tg mice express a soluble form of OVA in the serum under KJ1-26ϩCD4ϩCD25Ϫ (anergic) and CD25ϩ (Treg) T cells from DO11 ϫ control of the metallothionein promoter I and were generated by cloning sOVA Tg and KJ1-26ϩCD4ϩCD25Ϫ cells from DO11 mice (naive) were the OVA cDNA into the metallothionein control region expression sorted from lymph nodes. Activated cells were recovered as cycled KJ1– vector 2999 (kindly provided by R. D. Palmiter (University of Washington, 26ϩCD4ϩ cells from BALB/c mice that had been transferred with 5 ϫ 106 Seattle, WA) (19) and injecting the construct into blastocysts from FVB CFSE-labeled DO11 cells and immunized with 200 ␮g of OVA in IFA 4 mice. Founders were screened for OVA expression by Southern blotting days before sorting. Cells from up to four mice were pooled for each and ELISA (anti-OVA; Research Diagnostics), and one founder expressing ϳ individual sample per group. Total RNA was isolated using the Absolutely 20 ng/ml OVA in the serum was selected. Mice were subsequently typed RNA RT-PCR Miniprep from Stratagene. To avoid contamination with for the presence of OVA by PCR (forward primer, GCAATGCCTTTCA DNA, samples were treated with DNase (Ambion) before amplification. A GAGTGA; reverse primer, GCCCTAAATTCTTCAGAGACG). sOVA Tg total of 56.76 ng of RNA per sample was used for linear in vitro amplifi- mice have normal life expectancy. The mice were backcrossed onto the Ͼ cation using Superscript II (Invitrogen Life Techologies) for reverse tran- BALB/c background for 10 generations and crossed with DO11 TCR Tg scription and the MEGAscript T7 kit to generate cRNA (Ambion). The mice. Rat insulin promoter (RIP)-membrane-bound OVA (mOVA) Tg cRNA was used for a second round of amplification and finally labeled mice have been described (20, 21). For some experiments they were bred Downloaded from Ϫ/Ϫ with biotinylated ribonucleotides using the Enzo BioArray kit onto a Rag2 background. (Affymetrix). All mice were bred and maintained in our pathogen-free facility in ac- cordance with the guidelines of the Laboratory Animal Resource Center of Real-time RT-PCR the University of California, San Francisco. All experiments were con- ducted with the approval of the Committee on Animal Research of the Quantitative RT-PCR was performed using real-time fluorogenic PCR University of California, San Francisco. (TaqMan) on a PE Biosystems ABI Prism 7700 Sequence BioDetector according to the manufacturer’s instructions (PerkinElmer). Total RNA http://www.jimmunol.org/ Abs and flow cytometry was extracted as described above and reverse transcribed using Superscript CD4ϩ cells were stained with the clonotypic Ab KJ1-26 (Caltag Labora- II kit for RT-PCR (Invitrogen Life Technologies). No linear amplification was performed for RT-PCR. Primer and probe sequences for IL-4 and IFN- tories), anti-CD4 (GK1.5, H129.19, and RM4-5), and anti-CD25 (PC61, ␥ 7AD). All Abs were obtained from BD Pharmingen unless otherwise , and hypoxanthine phosphoribosyltransferase (HPRT) were used as pub- stated. Abs were used as FITC, PE, PE-Cy7, PE-Texas Red, allophyco- lished (23). cyanin, or PerCp conjugates. Fc Block (anti-CD16/CD32; BD Biosciences) GeneChip hybridization and analysis was added before staining. Flow cytometric analyses were done on a FACSCalibur with CellQuest software (both from BD Biosciences) or on Ten micrograms of cRNA was used to hybridize each MOE430A Gene- a CyAn analyzer (DakoCytomation). Cells were sorted with a MoFlo cell Chip array (Affymetrix). Hybridization was performed in a GeneChip Hy- sorter (DakoCytomation). For intracellular cytokine staining, transferred bridization Oven 640 (Affymetrix) for 16 h at 45°C. The arrays were DO11 T cells recovered from the peripheral or pancreatic lymph nodes of washed and stained on a GeneChip Fluidics Station 450 (Affymetrix), and by guest on September 29, 2021 Ϫ Ϫ RIP-mOVA Rag / recipients were restimulated on mitomycin C-treated scanned with a GeneChip Scanner 3000 (Affymetrix). Intensities of Perfect BALB/c splenocytes for 14 h in the presence of 1 ␮g/ml OVA-peptide. Match and Mismatch probes were generated by GeneChip operating soft- Brefeldin A (Epicentre Biotechnologies) was added (10 ␮g/ml) for the last ware 1.2 (Affymetrix). Gene expression was adjusted with the robust multi- 2 h of stimulation. Cells were stained for the intracellular cytokines IL-2 array average method using Bioconductor release 1.3. Robust multiarray and IFN-␥ and analyzed by flow cytometry. Staining with appropriate iso- average expression measurements of all probe sets were analyzed with type controls showed no detectable differences between experimental one-way ANOVA using version R 1.8.1. Differentially expressed probe groups. sets between groups were considered significant if their p values were Ͻ0.01 and their fold-change Ͼ2.0. For hierarchical clustering, differen- Cell preparations, purification, and adoptive transfer tially expressed probe sets were combined and clustered using GeneSpring CD4ϩKJ1-26ϩCD25Ϫ and CD25ϩ cells from DO11, DO11 ϫ sOVA Tg, 6.0 (Silicon Genetics). For identical genes, probe sets displaying the great- or DO11 ϫ RIP-mOVA Tg mice were recovered by cell sorting from est fold change are shown. lymph nodes or spleen. For adoptive transfer into wild-type recipients, 2ϩ sorted cells were labeled with 5 ␮M CFSE (Invitrogen Life Technologies) Ca flux analysis by flow cytometry ϫ 6 ϩ at 10 10 cells/ml for 10 min at 37°C and washed before injection. A For Ca2 analysis, lymph node cells or splenocytes were enriched for ϫ 6 ϩ ϩ Ϫ Ͼ ϩ ϩ ϩ total of 0.5–1 10 CD4 KJ1-26 CD25 cells (purity 95%) were CD4 cells by depletion of B220 and CD8 cells (Dynal), stained for Ϫ/Ϫ adoptively transferred by tail vein injection. In experiments using Rag CD4 and KJ1-26, and labeled with 5 ␮g/ml indo-1 acetoxymethyl ester ϫ 5 ϩ ϩ Ϫ ϩ ϩ recipients, a total of 2 10 CD4 KJ1-26 CD25 or CD25 cells were (Invitrogen Life Technologies). Ca2 flux was induced by adding soluble ␮ adoptively transferred, and the mice were immunized with 200 gofOVA anti-CD3 (2C11; BD Pharmingen). Soluble anti-CD3 was used at 5 ␮g/ml protein in IFA s.c. the following day. Activated DO11 cells for gene array in experiments with nonpurified DO11 ϫ sOVA Tg T cells and at 40 ␮g/ml were recovered by cell sorting from BALB/c or sOVA Tg mice that had for adoptively transferred DO11 cells because of the higher cell density per ϫ 6 ϩ been adoptively transferred with 5 10 CFSE-labeled, purified CD4 sample and low frequency of DO11 cells. Ionomycin (0.5 - 1 ␮g/ml; Sig- cells from the spleen and the lymph nodes of DO11 mice using magnetic ma-Aldrich) was used as a positive control. Ca2ϩ flux was analyzed on a ␮ beads (Dynal) and were immunized with 200 g of OVA protein in IFA CyAn flow cytometer with an Enterprise 621 UV laser by gating on adop- (Invitrogen Life Technologies) 4 days before sorting. tively transferred, CFSE-labeled KJ1-26ϩCD4ϩ, KJ1-26ϩCD4ϩCD25Ϫ T In vitro proliferation and cytokine assays cells in the TCR Tg mice, respectively. The ratio of emission at 400:40 and 450:50 nm over time is shown. Five thousand to 25,000 sorted KJ1-26ϩCD4ϩCD25Ϫ T cells from DO11 ϫ sOVA Tg and DO11 mice or sOVA Tg and BALB/c transfer Results ϫ 5 recipients were cultured with 0.25 10 mitomycin C-treated BALB/c Encounter with self-Ag leads to T cell deletion, anergy, and the splenocytes in 200 ␮l of RPMI 1640 medium that contained 10% FCS in 96-well plates (Costar). Cells were stimulated with 0–1 ␮g/ml OVA. development of regulatory T cells 3 [ H]Thymidine (1 ␮Ci/well) was added during the last 16 h of culture, and The initial experiments were designed to establish the fate of incorporation was measured by scintillation counting after 72 h of culture. Unless otherwise indicated, supernatants were collected after 48 h, and DO11 T cells that encountered OVA throughout development as a levels of IL-2 or IFN-␥ were assayed by ELISA as previously described ubiquitously secreted “self-Ag” expressed systemically. To do (22). For coculture assays 25,000 sorted KJ1-26ϩCD4ϩCD25Ϫ T cells this, we have generated Tg mice that express a soluble form of The Journal of Immunology 6475

OVA and crossed them with DO11 TCR Tg mice. To compare this form of systemic tolerance with tolerance to the self-Ag restricted to tissues, RIP-mOVA mice expressing membrane OVA in islet ␤ cells were also crossed with DO11 mice. As expected, self-Ag recognition during development results in the deletion of a sub- stantial proportion of the T cells in the thymus, and this is reflected in reduced numbers of DO11 cells in peripheral tissues (Fig. 1A). Thymic deletion is greater in DO11 ϫ sOVA Tg mice than in DO11 ϫ RIP-mOVA Tg mice, perhaps because the secreted Ag is more abundantly expressed in the thymus of sOVA Tg animals. Within the surviving DO11 cells ϳ30% expressed high levels of CD25, and the majority of these also expressed high levels of CD62L (Fig. 1, A and B), a phenotype that is characteristic of Treg (24). The same pattern of partial T cell deletion and development of CD25ϩ cells is seen when the DO11 ϫ sOVA Tg is on a RagϪ/Ϫ background, ensuring a monoclonal T cell population (Fig. 1C). Thus, both the soluble self-Ag and the membrane-asso-

ciated Ag induce deletion and the development of a population that Downloaded from is phenotypically similar to Tregs. In the next series of experiments, we examined the functional capabilities of the CD25Ϫ and CD25ϩ DO11 T cells purified from the lymphoid organs of the DO11 ϫ sOVA Tg mice. We postu- lated that the surviving CD25Ϫ T cells from the DO11 ϫ sOVA

Tgs are unresponsive (anergic) to self-Ag, because the mice do not http://www.jimmunol.org/ show any manifestations of autoimmunity. To test this hypothesis, CD25Ϫ DO11 cells were purified from these mice by cell sorting, labeled with CFSE, and transferred into sOVA Tg recipients. As shown in Fig. 2A, naive DO11 T cells (from TCR Tgs without the Ag) undergo rapid cycling after an encounter with the self-Ag in vivo following adoptive transfer (in separate studies we have found that this initial proliferation is followed by functional unre-

sponsiveness and cell death; Ref. 18). In contrast, DO11 T cells by guest on September 29, 2021 from the DO11 ϫ sOVA Tgs respond much less to the self-Ag. We next examined the responses of these T cells to stimulation with the Ag ex vivo. As shown in Fig. 2B, the CD25Ϫ DO11 cells from the DO11 ϫ sOVA mice are markedly hyporesponsive to stimulation with Ag, although their proliferative responses are re- stored at high Ag concentrations. Thus, these experiments showed that encounter with self-Ag makes CD25Ϫ DO11 T cells function- ally unresponsive and that this form of anergy can be reversed by stimulation at high Ag concentrations. Under the same conditions, CD25ϩ cells from the DO11 ϫ sOVA Tg mice are completely unresponsive to Ag in vitro, as has been demonstrated previously in many similar systems (16, 25, 26).

Anergic CD25Ϫ cells do not have suppressive activity In the next series of experiments, we asked whether the anergic CD25Ϫ populations had any suppressive activity. In coculture as- ϩ FIGURE 1. Thymic deletion and development of CD25 T cells in re- says, CD25ϩ cells from DO11 ϫ sOVA Tg mice profoundly sup- sponse to self-Ag. A, Lymph nodes (LN) and thymus (THY) of DO11, press the responses of normal DO11 T cells in a dose-dependent DO11 ϫ sOVA Tg, and DO11 ϫ RIP-mOVA Tg mice were harvested, total ϩ numbers of DO11 cells were measured by cell counting, and the percentage of manner (Fig. 3). For comparison, we included CD25 T cells from ϩ ϩ ϫ KJ1-26 CD4 cells was determined by flow cytometry. Cells were stained for DO11 RIP-mOVA Tg mice, which have previously been shown KJ1-26, CD4, and CD25 to determine the percentage of CD25ϩ cells. Each to have potent suppressive activity (26). The suppression is Ag symbol represents one individual mouse. B, FACS plots show the expression dependent, because CD25ϩCD4ϩ cells from normal BALB/c an- ϩ ϩ of CD25 and CD62L in KJ1-26 CD4 cells from lymph nodes. Data for a imals do not inhibit the responses of normal DO11 cells to OVA. representative mouse from four experiments with at least two mice per group In contrast to the CD25ϩ cells, CD25Ϫ T cells from either are shown. C, Thymi from DO11 RagϪ/Ϫ or DO11 ϫ sOVA Tg RagϪ/Ϫ mice DO11 ϫ sOVA or DO11 ϫ RIP-mOVA Tg mice are not suppres- were harvested and stained as described in A. Lower graphs are gated on ϩ sive (Fig. 3). The conclusion of these experiments is that self-Ag CD4 single positive thymocytes. Numbers refer to the percentage of gated ϩ cells in the upper right quadrant. Data from one representative mouse from encounter induces two populations of T cells: CD25 regulatory Ϫ two individual experiments with two mice each are shown. cells and CD25 anergic, but not regulatory, cells. 6476 ANERGIC AND REGULATORY T LYMPHOCYTES

FIGURE 2. CD4ϩCD25Ϫ T cells from DO11 ϫ sOVA Tg mice are hyporesponsive to Ag in vivo and in vitro. A, KJ1-26ϩCD4ϩ CD25Ϫ cells from DO11 or DO11 ϫ sOVA Tg mice were isolated by cell sorting, la- beled with CFSE, and transferred into either BALB/c wild-type or sOVA Tg mice. The CFSE dilution profile was analyzed 4 days after transfer. Numbers refer to the ratio of divided to undivided CD4ϩKJ1-26ϩ cells (proliferation index). One representative

experiment of three is shown. B, KJ1– Downloaded from 26ϩCD4ϩCD25Ϫ and CD25ϩ cells from DO11 ϫ sOVA Tg and DO11 ϫ RIP- mOVA Tg mice or KJ1-26ϩCD4ϩCD25Ϫ cells from DO11 mice were isolated by cell sorting and cultured with Ag plus APCs. Empty squares represent KJ1-26ϩCD4ϩ Ϫ CD25 cells from DO11 control mice in all http://www.jimmunol.org/ panels. [3H]Thymidine incorporation was as- sayed on day 3, and supernatants were har- vested for cytokine ELISA from the cultures on day 2. Data from one of three experiments are shown. by guest on September 29, 2021

Gene expression profiles of anergic and regulatory T cells Cd25, Ctla4, and a number of inhibitory cytokines, chemokines, Because anergic and regulatory T cells that are chronically ex- and chemokine receptors (Table I and supplemental data Table I) posed to self-Ag show quite different functional response poten- (29–31). Similarly, a distinct gene expression profile can be found tials, we asked whether their functions are reflected in the patterns in activated cells (Fig. 4, A and B). The expression level of 149 of genes expressed in these cells. We isolated CD25Ϫ (anergic) genes is uniquely regulated in activated cells, many of which play and CD25ϩ (regulatory) DO11 cells from the DO11 ϫ sOVA Tg a role in cell division and cell cycle control (Table I and supple- mice as well as DO11 cells that had been transferred into BALB/c mental data Table I). recipients and activated by immunization. We compared their gene In contrast, anergic T cells do not express a clear molecular expression profiles with naive DO11 CD25Ϫ T cells using Af- signature (Fig. 4, A and B). Only 24 individual genes are uniquely fymetrix gene chips displaying 22,000 probe sets. At the statistical regulated in anergic cells at a significant level when compared with thresholds described in Materials and Methods, we found a total of naive. Many of these have been described as aiding in peripheral 511 individual genes that are differentially up- or down-regulated deletion (e.g., FasL) or maybe altering the strength of signal trans- at least 2-fold in any of the anergic, activated, or regulatory T cell duction, such as Mapk12 (32, 33). Importantly, none of the genes population when compared with naive cells. Table I shows a list of that have been described as being responsible for regulatory T cell selected genes. The complete listing of differentially regulated function, like Foxp3, are expressed in anergic cells. It is also note- genes corresponding to the hierarchical clustering shown in Fig. 4 worthy that several genes that are up-regulated in activated T cells are can be found in the supplemental data Table I. also altered similarly or to a lower extent in anergic cells e.g., mem- bers of the TNF superfamily such as RANKL (Tnfsf11)orLIGHT The hierarchical clustering of targets expressed in the various 6 cell populations makes several important points (Fig. 4, A and B). (Tnfsf14) (supplemental data Table I). Surprisingly, some of the ␥ First, regulatory T cells have a distinct molecular signature that is genes expressed in anergic cells (such as IFN- and IL-4) are thought completely different from that of activated or anergic T cells. We to be characteristic of activated and effector T cells. These findings found 123 individual genes that were exclusively regulated in Treg suggest that anergy may reflect a partial activation phenotype. (Fig. 4B). Many of the genes that are uniquely regulated in Treg have been previously described (27, 28); these include Foxp3, 6 The online version of this article contains supplemental material. The Journal of Immunology 6477

FIGURE 3. CD4ϩCD25Ϫ T cells from DO11 ϫ sOVA Tg mice are an- ergic but not suppressive. KJ1– 26ϩCD4ϩCD25Ϫ or CD25ϩ cells from DO11 ϫ sOVA Tg mice and DO11 ϫ RIP-mOVA Tg mice were isolated by cell sorting and cocul- tured with KJ1-26ϩCD4ϩCD25Ϫ cells from DO11 mice with Ag plus APCs. The number of KJ1-26ϩCD4ϩ CD25Ϫ responder cells from DO11 mice remained constant. The number indicates the input ratio of KJ1– 26ϩCD4ϩCD25Ϫ or CD25ϩ to KJ1– 26ϩCD4ϩCD25Ϫ responder cells. CD4ϩCD25ϩ and CD4ϩCD25Ϫ cells Downloaded from from BALB/c mice were used as con- trols. [3H]Thymidine incorporation was assayed on day 3, and superna- tants were harvested for ELISA on day 2 of culture. http://www.jimmunol.org/

Several genes are regulated in the same fashion in activated, genic stimulus. To test this hypothesis, we used an adoptive regulatory, and anergic T cells (85 genes). These genes may be transfer model in which CD4-purified DO11 T cells were trans- markers of Ag recognition (Table I and supplemental data Table I). ferred into sOVA Tg or WT BALB/c recipients (18). After 5 Among these genes are the chemokine receptors CCR9 and days, the DO11 cells were isolated from the lymph node and CXCR3, known to be expressed on intestinal homing lymphocytes spleen by cell sorting and restimulated on Ag ex vivo. Fig. 5A by guest on September 29, 2021 and activated Th1 cells, respectively (34–36), and the costimula- shows that the DO11 cells that had encountered the Ag were tory molecule ICOS (37, 38). In contrast, genes that are shared defective in proliferation and IL-2 production and did not show between regulatory and anergic but not activated cells most likely any detectable IFN-␥ production (data not shown) upon re- reflect exposure to self-Ag rather than regulatory potency and in- stimulation with Ag and splenocytes ex vivo. Thus, they are clude surface receptors that have been associated with signaling considered hyporesponsive and are functionally tolerant. To de- modification in anergic cells and regulatory T cells (e.g., CD5, termine the expression of IFN-␥ and IL-4 mRNA, quantitative GITR, CD38, and neuropilin) (Table I) (28, 39–44). real-time PCR was performed for DO11 cells that had encoun- Therefore, within the groups of genes that are regulated in the tered the Ag in sOVA Tg mice or naive DO11 cells. Fig. 5B same fashion in anergic and activated cells, anergic and Treg cells, shows that self-Ag recognition leads to strong expression of or in all three groups, the great majority of the genes is regulated to the IFN-␥ and IL-4 mRNA 4 days after transfer. Thus, anergy in- same or greater degree in activated or regulatory cells when compared duction is accompanied by partial effector differentiation in with anergic cells (Fig. 4B and supplemental data Table I). This vivo, and the anergic cells seem to be poised to turn into ef- means that the magnitude of gene regulation in anergic cells, for the fector cells but are incapable of secreting the cytokine in a most part, lies between naive and regulatory or naive and activated tolerogenic environment. cells. The gene expression profile of anergic cells may therefore re- In view of these findings, we hypothesized that tolerogenic stim- flect partial activation or Ag recognition in general, but failure to fully uli must use other signaling pathways in parallel that prevent ef- commit to either the effector or Treg lineage. fective cytokine production. To test this hypothesis, we isolated DO11 cells that had been transferred into sOVA Tg or BALB/c Anergy induction in vivo is associated with partial effector recipients and measured Ca2ϩ mobilization in response to TCR differentiation and proximal signaling defects cross-linking. As shown in Fig. 6A, DO11 cells that have encoun- The genes that are regulated in anergic cells, when compared tered the self-Ag show a defect in Ca2ϩ flux when compared with with naive T cells, appear to fall into two expression patterns. naive cells. Although cell cycling seems to further promote this Anergic cells express some genes such as IFN-␥ and IL-4, signaling block, defective Ca2ϩ mobilization can already be seen which are typically found in effector cells, and also express in cells that have not divided. Thus, the induction of anergy is other genes such as PD-1, which are shared with regulatory T simultaneously associated with the development of a proximal cells and thus most likely represent genes that are induced by signaling defect as well as with partial effector differentiation. A continuous exposure to tolerogenic Ag. We therefore hypothe- similar Ca2ϩ mobilization defect can be seen in CD25Ϫ DO11 cells sized that the induction of anergy is accompanied by simulta- from DO11 ϫ sOVA Tg mice (Fig. 6B). Therefore, it should be neous differentiation along an aborted effector pathway and a possible for anergic cells to acquire full effector function if the tolerogenic pathway, with decreased responsiveness to an anti- antigenic stimulus is strong enough to overcome the signaling block. 6478 ANERGIC AND REGULATORY T LYMPHOCYTES

Table I. Selected gene expression of anergic, activated, and regulatory T cells compared with naivea

GenBank No. Gene Name Anergic/Naive Treg/Naive Active/Naive Description

Anergic specific NM_013653 Ccl5 4.6 Chemokine (C-C motif) ligand 5 K00083 Ifng 4.0 IFN-␥ BC021640 Mapk12 2.6 MAPK-12 (ERK-6) NM_019492 Rgs3 2.5 Regulator of G protein signaling 3 BC010198 Sh3bp2 2.1 SH3 domain-binding protein 2 NM_007913 Egr1 4.2 Early growth response 1 NM_009452 Tnfsf4 4.5 OX40L, TNF (ligand) superfamily member 4 NM_010177 Tnfsf6 3.4 FasL, TNF (ligand) superfamily member 6 Overlap anergic- Treg NM_007650 Cd5 4.9 4.0 CD5 Ag (p56-62) BB256012 Cd38 4.1 5.3 CD38 Ag AK002673 Nrp 6.9 13.3 Neuropilin-1 AF229434 Tnfrsf18 2.6 7.1 TNFR superfamily member 18 (GITR) M59378 Tnfrsf1b 2.4 6.7 TNFR superfamily member 1b (TNFR2)

BM250782 Tnfrsf9 19.4 22.9 TNFR superfamily member 9 (4-1BB) Downloaded from Overlap anergic- activated NM_021283 Il4 3.0 2.4 IL-4 NM_021782 Il21 8.6 4.3 IL-21 NM_019418 Tnfsf14 4.1 4.0 TNF (ligand) superfamily member 14 (LIGHT) NM_011613 Tnfsf11 7.3 9.4 Tnf (ligand) superfamily member 11 (RANKL)

Overlap anergic- http://www.jimmunol.org/ activated-Treg BC008152 Casp1 3.6 3.5 4.7 1 BG070529 Casp3 3.3 4.4 5.3 Caspase 3, -related cysteine protease NM_007609 Casp4 6.7 5.4 8.4 Caspase 4, apoptosis-related cysteine protease NM_007611 Casp7 2.1 2.6 2.8 Caspase 7 NM_008798 Pdcd1 16.9 9.7 4.9 Programmed cell death 1 NM_021396 Pdcd1Ig2 3.2 3.4 2.3 Programmed cell death 1 ligand 2 NM_008368 Il2rb 3.7 9.8 4.5 IL-2R, ␤-chain NM_013674 Irf4 4.3 6.4 6.0 IFN regulatory factor 4

X06746 Egr2 9.3 6.6 5.2 Early growth response 2 by guest on September 29, 2021 AB023132 Icos 5.6 10.1 6.5 Inducible T cell costimulator NM_008479 Lag3 11.8 6.9 2.2 Lymphocyte activation gene 3 NM_013869 Tnfrsf19 4.8 2.1 3.8 TNFR superfamily member 19 (TRADE) NM_011659 Tnfrsf4 7.2 25.7 6.5 TNFR superfamily member 4 (OX40) Treg specific AF054581 Il2ra 18.3 IL-2R, ␣-chain NM_010548 Il10 2.8 IL-10 BM248342 Tgfbr1 4.9 TGF-␤ receptor 1 NM_009895 Cish 5.8 Cytokine-inducible SH2-containing protein AK007410 Gadd45g 3.8 Growth arrest and DNA damage-inducible 45 ␥ NM_007706 Socs2 16.8 Suppressor of cytokine signaling 2 BG064103 Traf1 4.0 TNFR-associated factor 1 NM_054039 Foxp3 21.3 Foxhead box P3 AF155372 Nfkb2 2.7 NF of ␬ light polypeptide gene enhancer in B cells 2 NM_009046 Relb 2.5 Avian reticuloendotheliosis viral (v-rel) oncogene-related B NM_009843 Ctla4 16.2 Cytotoxic T lymphocyte-associated protein 4 BC008626 Icam1 2.1 Intercellular adhesion molecule 1 ␣ NM_008399 Itgae 36.9 Integrin E, epithelium associated (Ag CD103) Activated specific NM_009862 Cdc451 3.5 Cell division cycle 45 homolog (Saccharomyces cerevisiae)-like NM_009870 Cdk4 2.2 Cyclin-dependent kinase 4 NM_020567 Gmnn 4.4 Geminin, DNA replication inhibitor NM_19499 Mad2l1 5.2 MAD2 (mitotic arrest deficient, homolog)-like 1 (yeast) BG075396 Tfdp1 3.4 Transcription factor Dp 1 AF065914 Il2 2.2 IL-2 X66083 Cd44 3.3 CD44 Ag

a GeneChip processing and statistical analysis were done as described in Materials and Methods and the Fig. 5 legend. Fold change values of anergic, activated, and regulatory T cells versus naive DO11 cells were calculated using robust multiarray average data and considered as significant with a cutoff of Ͼ2.0 and a p value of Ͻ0.01 determined by one-way ANOVA. Genes with a known function in T cells and those that were increased over naive were chosen. The complete listing of known differentially regulated genes can be found in supplemental data Table I. The Journal of Immunology 6479

FIGURE 4. Gene expression profiles of anergic, regulatory, and activated T cells compared with naive cells. KJ1-26ϩCD4ϩCD25Ϫ (anergic) and KJ1– 26ϩCD4ϩCD25ϩ (regulatory) cells from DO11 ϫ sOVA Tg mice, KJ1-26ϩCD4ϩCD25Ϫ (naive) cells from DO11 mice, and KJ1-26ϩCD4ϩ cells from trans- ferred and immunized BALB/c mice (activated; see Materials and Methods) were isolated by cell sorting. For each of the three replicates, RNA was processed from pooled cells of an individual experiment. Total RNA was extracted, amplified, labeled, and hybrid- ized on Affymetrix gene arrays. A, 591 differentially expressed probe sets in three group comparisons were combined for two-dimensional hierarchical clustering of significant genes. Hierarchical clustering was con- ducted on all differentially expressed probe sets from Downloaded from 12 GeneChip arrays. Red and blue colors, respec- tively, indicate high and low levels of expression. B, Solidly filled (black) bars represent the total number of individual genes that are exclusively regulated in the indicated group compared with expression level in na- ive cells. Mixed filled and open (white) bars show the total number of genes for which up- or down-regula- http://www.jimmunol.org/ tion is shared in the indicated groups compared with naive cells. Within these groups, the open fraction of the bar represents the number of genes that are regu- lated at least 2-fold more in anergic cells than in ac- tivated or regulatory T cells. by guest on September 29, 2021

CD25Ϫ anergic cells but not CD25ϩ regulatory cells can Discussion induce autoimmunity The studies in this paper were designed to analyze the properties of If responses of anergic cells can be restored with strong stimula- anergic and regulatory T cells induced in the same lymphocyte tion (Fig. 2B), it is possible that anergic cells retain the capacity for population by a known systemic Ag. By crossing a TCR Tg mouse causing harmful reactions. It is not known whether the same is true with a strain that expresses a circulating form of the Ag recognized of CD25ϩ Treg. by the TCR, we have generated both anergic and regulatory T cells To compare the pathogenic potential of these cell populations, with the same Ag specificity in the population that survives thymic we used an experimental system that we have described previously deletion. It is therefore possible to compare the functional re- (45) in which DO11 cells induce diabetes when transferred into sponses and biochemical properties of these cells in a way that is lymphopenic (RagϪ/Ϫ) RIP-mOVA recipients and immunized. not feasible in most experimental systems. CD25Ϫ or CD25ϩ DO11 cells were purified from DO11 ϫ sOVA In DO11 ϫ sOVA Tg mice, a large fraction of the DO11 cells Tg mice and transferred into RIP-mOVA RagϪ/Ϫ recipients. The is deleted in the thymus; among the survivors, ϳ30% express Ϫ recipient mice were immunized with OVA in adjuvant and fol- CD25 (Fig. 1). CD25 cells isolated from these mice are hypore- lowed for the development of diabetes. The CD25ϩ cells failed to sponsive to the same circulating form of OVA in vivo or to stim- induce diabetes. In contrast, all mice that had been transferred with ulation with the Ag ex vivo (Fig. 2), thus fulfilling the essential anergic CD25Ϫ cells and immunized developed diabetes within 3 criteria for T cell anergy. Interestingly, these anergic cells can be wk (Fig. 7A). DO11 T cells were isolated from these mice 4 wk induced to respond at high Ag concentrations (Fig. 2B). More sig- after transfer and analyzed for cytokine production by intracellular nificantly, the anergic cells are able to elicit pathologic reactions cytokine stains. Consistent with diabetes development, the CD25Ϫ against tissue (islet) OVA if the cells are removed from the sys- cells produced IL-2 and IFN-␥, whereas CD25ϩ regulatory T cells temic tolerogen and exposed to an immunogenic form of the Ag lacked cytokine production (Fig. 7B). Thus, immunization in a (Fig. 7). It has long been suspected that anergy can be broken by model of autoimmunity will make anergic cells autoaggressive, strong Ags or by an encounter with microbes, and this loss of the although the same anergic cells do not respond to systemic self-Ag tolerant state results in autoimmune reactions (4, 46–49). Our alone (Fig. 2A). In contrast to anergic cells, transfer into RIP- studies formally demonstrate that T cells that are chronically ex- mOVA RagϪ/Ϫ recipients followed by immunization does not posed to a circulating Ag are unresponsive to the same Ag in vivo break the unresponsiveness of regulatory T cells. but can react strongly to other forms of that Ag. Thus, anergy 6480 ANERGIC AND REGULATORY T LYMPHOCYTES

FIGURE 5. Recognition of tolerogenic Ag by na- ive T cells promotes expression of effector cytokine mRNA. CD4ϩ-purified DO11 T cells were transferred into WT BALB/c or sOVA Tg animals. Lymphocytes were harvested from peripheral lymph nodes on day 5 (A)orday4(B), and CD4ϩKJ1-26ϩ cells were iso- lated by cell sorting. A, Sorted CD4ϩKJ1-26ϩ cells were restimulated with APCs and OVA at the indi- cated doses. [3H]Thymidine incorporation was as- sayed on day 3, and IL-2 secretion by ELISA was measured on day 2. B, mRNA was isolated from sorted cells and analyzed for the expression of IFN-␥, IL-4, and HPRT by real-time fluorogenic RT-PCR.

The abbreviation n.d. represents the amounts of tran- Downloaded from script that are not detectable. Transcript abundance is represented as the ratio of cytokine to HPRT. Data are from one representative experiment of three. http://www.jimmunol.org/

appears to be a state of “desensitization” that is not permanent and the responses of normal DO11 cells in coculture assays (Fig. 3). In can be overcome by strong external stimuli (17, 50–52). this experimental situation, CD25Ϫ DO11 cells have no regulatory The CD25ϩ cells that develop in the DO11 ϫ sOVA Tg mice function. This is different from other double Tg models in which have the functional characteristics of Tregs because they inhibit CD25Ϫ cells also show suppressive function, and it may be related by guest on September 29, 2021

FIGURE 6. Anergic CD4ϩ cells have a cell cycle-independent defect in Ca2ϩ-mobilization. A, CFSE-labeled CD4ϩ T cells from DO11 mice were adoptively transferred into BALB/c (a) or sOVA Tg recipients on day 0. Splenocytes were recovered 3 days later, and the Ca2ϩ flux of KJ1-26ϩCD4ϩ cells gated on cells that had divided (c) or not divided (b) was measured after stimulation with anti-CD3 as described in Materials and Methods. The table summarizes data from day 3 (d3) to day 5 (d5) with two mice per group. B, Splenocytes from DO11 ϫ sOVA Tg mice or DO11 Tg mice were isolated, stained for KJ1-26, CD4, and CD25, and labeled with indo-1 dye. Ca2ϩ flux was measured in gated KJ1-26ϩCD4ϩCD25Ϫ cells after stimulation with soluble anti-CD3. The table summarizes the percentage of KJ1-26ϩCD4ϩCD25Ϫ cells that flux Ca2ϩ. Data from one of five experiments is shown. The Journal of Immunology 6481

FIGURE 7. Anergic cells can in- duce autoimmune diabetes. KJ1– 26ϩCD4ϩCD25Ϫ and KJ1-26ϩCD4ϩ CD25ϩ cells from DO11 ϫ sOVA Tg mice or KJ1-26ϩCD4ϩCD25Ϫ cells from DO11 mice were isolated by cell sorting, and 2 ϫ 105 cells were trans- ferred into RIP-mOVA RagϪ/Ϫ mice.

The recipients were immunized with Downloaded from OVA/IFA the following day. A, Blood glucose readings of recipient mice from week 0 to week 6 are shown. All mice transferred with CD25ϩ cells remained normoglycemic for Ͼ5 wk. Data are from one representative experiment of two. B, Peripheral and pancreatic lymph http://www.jimmunol.org/ nodes (LN) of recipients were harvested after 4 wk, restimulated with OVA and APCs, and stained for intracellular cy- tokines. The average percentage of cy- tokine-positive KJ1-26ϩCD4ϩ cells from two individual mice per group is shown. One representative experiment of two is shown. by guest on September 29, 2021

to the amount of the Ag or to the way the Ag is presented (16). As Activated cells present a unique gene expression profile that expected, the CD25ϩ Tregs are incapable of mediating pathologic distinguishes them from both anergic and regulatory T cells. In reactions. By this criterion, Tregs are fundamentally different from contrast, anergic cells do not have such a clear signature (Fig. 4). anergic T cells. We find that anergic cells regulate many genes in the same fashion To examine the molecular basis of T cell anergy, we have ini- as activated or regulatory T cells or both groups. However, the tiated a study of gene expression profiles in anergic T cells. An great majority of these genes are regulated at the same or higher important aspect of our studies is that, for the first time, we have magnitude in the activated or regulatory population as compared compared anergic and regulatory T cells expressing the same Ag with anergic cells. This finding suggests that anergic cells have receptor and induced by the same self-Ag. This analysis reveals responded to self-Ag but failed to become fully activated or to interesting differences between the two cell populations. develop into regulatory cells. It is interesting to note that many Regulatory cells have a clear and distinct molecular signature, as genes are regulated in a similar fashion in both anergic and is now widely accepted (27, 28). We found many of the genes that have been described and would be expected to be regulatory based regulatory T cells (Fig. 4 and supplemental Table I). These on various methods of detection, the most abundantly expressed of overlapping genes, such as CD5, GITR, CD38, and neuropilin, which are Cd25, Foxp3, and Ctla4 (29–31). Furthermore, it has may reflect Ag encounter in a tolerogenic rather than immuno- recently been shown that Treg efficiency can be triggered directly genic fashion and may be necessary for preventing the response via TLR stimulation (53). The NF-␬B-related genes (RelB, Traf1, of T cells, e.g., by varying activation thresholds, but do not and NF␬B2) that are up-regulated in Treg may be important in confer full commitment to the regulatory T cell lineage. CD5 mediating this effect. has been suggested to play a role in down-regulation of TCR 6482 ANERGIC AND REGULATORY T LYMPHOCYTES responses by recruitment of SHP-1 protein phosphatase (54), 13. Lanoue, A., C. Bona, H. von Boehmer, and A. Sarukhan. 1997. Conditions that induce tolerance in mature CD4ϩ T cells. J. Exp. Med. 185: 405–414. and CD5 protein surface expression in our model correlates 14. Adler, A. J., C. T. Huang, G. S. Yochum, D. W. Marsh, and D. M. Pardoll. 2000. with RNA levels (data not shown). In vivo CD4ϩ T cell tolerance induction versus priming is independent of the rate How a tolerogenic stimulus is different from an immunogenic and number of cell divisions. J. Immunol. 164: 649–655. 15. Jordan, M. S., M. P. Riley, H. von Boehmer, and A. J. Caton. 2000. Anergy and stimulus and how these stimuli result in strikingly different func- suppression regulate CD4(ϩ) T cell responses to a self peptide. Eur. J. Immunol. tional consequences are fundamental questions. Our results sug- 30: 136–144. 16. Apostolou, I., A. Sarukhan, L. Klein, and H. von Boehmer. 2002. Origin of gest that lymphocyte activation and anergy are associated with regulatory T cells with known specificity for antigen. Nat. Immunol. 3: 756–763. surprisingly overlapping cellular responses. For instance, even an- 17. Singh, N. J., and R. H. Schwartz. 2003. The strength of persistent antigenic ϩ ergic T cells express abundant transcripts for cytokines thought to stimulation modulates adaptive tolerance in peripheral CD4 T cells. J. Exp. ␥ Med. 198: 1107–1117. be typical of effector responses (IFN- and IL-4). The reason why 18. Lohr, J., B. Knoechel, E. C. Kahn, and A. K. Abbas. 2004. Role of B7 in T cell anergic cells do not continuously produce these cytokines when tolerance. J. Immunol. 173: 5028–5035. exposed to a self-Ag is because anergy is associated with a prox- 19. Palmiter, R. D., E. P. Sandgren, D. M. Koeller, and R. L. Brinster. 1993. Distal regulatory elements from the mouse metallothionein locus stimulate gene expres- imal signaling block in response to the Ag. This phenomenon has sion in transgenic mice. Mol. Cell. Biol. 13: 5266–5275. been called “tuning” of the activation threshold of T cells (17). 20. Kurts, C., H. Kosaka, F. R. Carbone, J. F. Miller, and W. R. Heath. 1997. Class I-restricted cross-presentation of exogenous self-antigens leads to deletion of au- Clearly, however, anergic cells retain their capacity to develop into toreactive CD8(ϩ) T cells. J. Exp. Med. 186: 239–245. effector cells, even effector cells capable of causing disease 21. Lohr, J., B. Knoechel, S. Jiang, A. H. Sharpe, and A. K. Abbas. 2003. The (Fig. 7). inhibitory function of B7 costimulators in T cell responses to foreign and self- antigens. Nat. Immunol. 4: 664–669. The situation appears to be fundamentally different in regulatory 22. Howland, K. C., L. J. Ausubel, C. A. London, and A. K. Abbas. 2000. The roles Downloaded from T cells, which show functional responses reflected in the patterns of CD28 and CD40 ligand in T cell activation and tolerance. J. Immunol. 164: 4465–4470. of gene expression that are very different from those in anergic- 23. Grogan, J. L., M. Mohrs, B. Harmon, D. A. Lacy, J. W. Sedat, and cells. These differences suggest that anergic and regulatory cells do R. M. Locksley. 2001. Early transcription and silencing of cytokine genes un- not share a common lineage or developmental pathway. derlie polarization of T helper cell subsets. Immunity 14: 205–215. 24. Sakaguchi, S. 2000. Regulatory T cells: key controllers of immunologic self- In summary, by using a simple Tg experimental system with a tolerance. Cell 101: 455–458. true self-Ag, our results demonstrate that anergy as a mechanism 25. Kawahata, K., Y. Misaki, M. Yamauchi, S. Tsunekawa, K. Setoguchi, of tolerance is a fundamentally different process from the devel- J. Miyazaki, and K. Yamamoto. 2002. Generation of CD4(ϩ)CD25(ϩ) regula- http://www.jimmunol.org/ tory T cells from autoreactive T cells simultaneously with their negative selection opment of regulatory T cells. It is important to determine whether in the thymus and from nonautoreactive T cells by endogenous TCR expression. either or both of these control pathways are triggered by exposure J. Immunol. 168: 4399–4405. 26. Walker, L. S., A. Chodos, M. Eggena, H. Dooms, and A. K. Abbas. 2003. An- to different forms of Ags, including self-Ags and chronic microbial tigen-dependent proliferation of CD4ϩCD25ϩ regulatory T cells in vivo. J. Exp. infections. Gene expression profiling may be an approach for com- Med. 198: 249–258. paring the relative development of these control mechanisms and 27. Gavin, M. A., S. R. Clarke, E. Negrou, A. Gallegos, and A. Rudensky. 2002. Homeostasis and anergy of CD4(ϩ)CD25(ϩ) suppressor T cells in vivo. Nat. how they can be manipulated. Immunol. 3: 33–41. 28. McHugh, R. S., M. J. Whitters, C. A. Piccirillo, D. A. Young, E. M. Shevach, Acknowledgments M. Collins, and M. C. Byrne. 2002. CD4(ϩ)CD25(ϩ) immunoregulatory T cells: by guest on September 29, 2021 We are indebted to S. Jiang and C. McArthur for expert cell sorting and gene expression analysis reveals a functional role for the glucocorticoid-induced TNF receptor. Immunity 16: 311–323. C. Benitez for mouse typing. We thank the members of the Abbas Labo- 29. Hori, S., T. Nomura, and S. Sakaguchi. 2003. Control of regulatory T cell de- ratory and Bluestone Laboratory for helpful discussion. velopment by the transcription factor Foxp3. Science 299: 1057–1061. 30. Fontenot, J. D., M. A. Gavin, and A. Y. Rudensky. 2003. Foxp3 programs the Disclosures development and function of CD4ϩCD25ϩ regulatory T cells. Nat. Immunol. 4: 330–336. The authors have no financial conflict of interest. 31. Khattri, R., T. Cox, S. A. Yasayko, and F. Ramsdell. 2003. An essential role for Scurfin in CD4ϩCD25ϩ T regulatory cells. Nat. Immunol. 4: 337–342. References 32. Chiariello, M., M. J. Marinissen, and J. S. Gutkind. 2000. Multiple mitogen- 1. Rocha, B., and H. von Boehmer. 1991. Peripheral selection of the T cell reper- activated protein kinase signaling pathways connect the cot oncoprotein to the toire. Science 251: 1225–1228. c-jun promoter and to cellular transformation. Mol. Cell. Biol. 20: 1747–1758. 2. Maloy, K. J., and F. Powrie. 2001. Regulatory T cells in the control of immune 33. Macian, F., F. Garcia-Cozar, S. H. Im, H. F. Horton, M. C. Byrne, and A. Rao. pathology. Nat. Immunol. 2: 816–822. 2002. Transcriptional mechanisms underlying lymphocyte tolerance. Cell 109: 3. Schwartz, R. H. 2003. T cell anergy. Annu. Rev. Immunol. 21: 305–334. 719–731. 4. Walker, L. S., and A. K. Abbas. 2002. The enemy within: keeping self-reactive 34. Zabel, B. A., W. W. Agace, J. J. Campbell, H. M. Heath, D. Parent, A. I. Roberts, T cells at bay in the periphery. Nat. Rev. Immunol. 2: 11–19. E. C. Ebert, N. Kassam, S. Qin, M. Zovko, et al. 1999. Human G protein-coupled 5. Shevach, E. M. 2002. CD4ϩCD25ϩ suppressor T cells: more questions than receptor GPR-9–6/CC 9 is selectively expressed on intesti- answers. Nat. Rev. Immunol. 2: 389–400. nal homing T lymphocytes, mucosal lymphocytes, and thymocytes and is re- 6. Sakaguchi, S., N. Sakaguchi, J. Shimizu, S. Yamazaki, T. Sakihama, M. Itoh, quired for thymus-expressed chemokine-mediated chemotaxis. J. Exp. Med. 190: Y. Kuniyasu, T. Nomura, M. Toda, and T. Takahashi. 2001. Immunologic tol- 1241–1256. erance maintained by CD25ϩCD4ϩ regulatory T cells: their common role in 35. Zaballos, A., J. Gutierrez, R. Varona, C. Ardavin, and G. Marquez. 1999. Cutting controlling autoimmunity, tumor immunity, and transplantation tolerance. Immu- edge: identification of the orphan chemokine receptor GPR-9–6 as CCR9, the nol. Rev. 182: 18–32. receptor for the chemokine TECK. J. Immunol. 162: 5671–5675. 7. Marrack, P., and J. Kappler. 2004. Control of T cell viability. Annu. Rev. Immu- 36. Rabin, R. L., M. A. Alston, J. C. Sircus, B. Knollmann-Ritschel, C. Moratz, nol. 22: 765–787. D. Ngo, and J. M. Farber. 2003. CXCR3 is induced early on the pathway of ϩ 8. Schwartz, R. H. 1996. Models of T cell anergy: is there a common molecular CD4 T cell differentiation and bridges central and peripheral functions. J. Im- mechanism? J. Exp. Med. 184: 1–8. munol. 171: 2812–2824. 9. Mueller, D. L., M. K. Jenkins, L. Chiodetti, and R. H. Schwartz. 1990. An in- 37. McAdam, A. J., T. T. Chang, A. E. Lumelsky, E. A. Greenfield, V. A. Boussiotis, tracellular calcium increase and protein kinase C activation fail to initiate T cell J. S. Duke-Cohan, T. Chernova, N. Malenkovich, C. Jabs, V. K. Kuchroo, et al. proliferation in the absence of a costimulatory signal. J. Immunol. 144: 2000. Mouse inducible costimulatory molecule (ICOS) expression is enhanced by 3701–3709. CD28 costimulation and regulates differentiation of CD4ϩ T cells. J. Immunol. 10. Jenkins, M. K., C. A. Chen, G. Jung, D. L. Mueller, and R. H. Schwartz. 1990. 165: 5035–5040. Inhibition of antigen-specific proliferation of type 1 murine T cell clones after 38. Coyle, A. J., S. Lehar, C. Lloyd, J. Tian, T. Delaney, S. Manning, T. Nguyen, stimulation with immobilized anti-CD3 monoclonal antibody. J. Immunol. 144: T. Burwell, H. Schneider, J. A. Gonzalo, et al. 2000. The CD28-related molecule 16–22. ICOS is required for effective T cell-dependent immune responses. Immunity 13: 11. Pape, K. A., R. Merica, A. Mondino, A. Khoruts, and M. K. Jenkins. 1998. Direct 95–105. evidence that functionally impaired CD4ϩ T cells persist in vivo following in- 39. Lechner, O., J. Lauber, A. Franzke, A. Sarukhan, H. von Boehmer, and J. Buer. duction of peripheral tolerance. J. Immunol. 160: 4719–4729. 2001. Fingerprints of anergic T cells. Curr. Biol. 11: 587–595. 12. Perez, V. L., L. Van Parijs, A. Biuckians, X. X. Zheng, T. B. Strom, and 40. Shimizu, J., S. Yamazaki, T. Takahashi, Y. Ishida, and S. Sakaguchi. 2002. Stim- A. K. Abbas. 1997. Induction of peripheral T cell tolerance in vivo requires ulation of CD25(ϩ)CD4(ϩ) regulatory T cells through GITR breaks immuno- CTLA-4 engagement. Immunity 6: 411–417. logical self-tolerance. Nat. Immunol. 3: 135–142. The Journal of Immunology 6483

41. Li, R., and D. M. Page. 2001. Requirement for a complex array of costimulators 48. Benoist, C., and D. Mathis. 2001. Autoimmunity provoked by infection: how in the negative selection of autoreactive thymocytes in vivo. J. Immunol. 166: good is the case for T cell epitope mimicry? Nat. Immunol. 2: 797–801. 6050–6056. 49. Stefanova, I., J. R. Dorfman, M. Tsukamoto, and R. N. Germain. 2003. On the 42. Page, D. M. 1999. Cutting edge: thymic selection and autoreactivity are regulated role of self-recognition in T cell responses to foreign antigen. Immunol. Rev. 191: by multiple coreceptors involved in T cell activation. J. Immunol. 163: 97–106. 3577–3581. 50. Adelstein, S., H. Pritchard-Briscoe, T. A. Anderson, J. Crosbie, G. Gammon, 43. Bruder, D., M. Probst-Kepper, A. M. Westendorf, R. Geffers, S. Beissert, R. H. Loblay, A. Basten, and C. C. Goodnow. 1991. Induction of self-tolerance K. Loser, H. von Boehmer, J. Buer, and W. Hansen. 2004. Neuropilin-1: a surface in T cells but not B cells of transgenic mice expressing little self antigen. Science marker of regulatory T cells. Eur. J. Immunol. 34: 623–630. 251: 1223–1225. 44. Read, S., S. Mauze, C. Asseman, A. Bean, R. Coffman, and F. Powrie. 1998. ϩ 51. Tanchot, C., D. L. Barber, L. Chiodetti, and R. H. Schwartz. 2001. Adaptive CD38ϩ CD45RB(low) CD4 T cells: a population of T cells with immune ϩ tolerance of CD4 T cells in vivo: multiple thresholds in response to a constant regulatory activities in vitro. Eur. J. Immunol. 28: 3435–3447. level of antigen presentation. J. Immunol. 167: 2030–2039. 45. Eggena, M. P., L. S. Walker, V. Nagabhushanam, L. Barron, A. Chodos, and A. K. Abbas. 2004. Cooperative roles of CTLA-4 and regulatory T cells in tol- 52. Grossman, Z., and W. E. Paul. 2001. Autoreactivity, dynamic tuning, and selec- erance to an islet cell antigen. J. Exp. Med. 199: 1725–1730. tivity. Curr. Opin. Immunol. 13: 687–698. 46. Sevilla, N., D. Homann, M. von Herrath, F. Rodriguez, S. Harkins, J. L. Whitton, 53. Caramalho, I., T. Lopes-Carvalho, D. Ostler, S. Zelenay, M. Haury, and and M. B. Oldstone. 2000. Virus-induced diabetes in a transgenic model: role of J. Demengeot. 2003. Regulatory T cells selectively express Toll-like receptors cross-reacting viruses and quantitation of effector T cells needed to cause disease. and are activated by lipopolysaccharide. J. Exp. Med. 197: 403–411. J. Virol. 74: 3284–3292. 54. Perez-Villar, J. J., G. S. Whitney, M. A. Bowen, D. H. Hewgill, A. A. Aruffo, and 47. Horwitz, M. S., A. Ilic, C. Fine, E. Rodriguez, and N. Sarvetnick. 2002. Presented S. B. Kanner. 1999. CD5 negatively regulates the T-cell antigen receptor signal antigen from damaged pancreatic ␤ cells activates autoreactive T cells in virus- transduction pathway: involvement of SH2-containing phosphotyrosine phospha- mediated autoimmune diabetes. J. Clin. Invest. 109: 79–87. tase SHP-1. Mol. Cell. Biol. 19: 2903–2912. Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021 GenBank/RefSeq Gene Name Anergic/Naïve Treg/Naïve Act./Naïve Description BF681826 Angptl2 7.7 4.2 2.7 angiopoietin-like 2 NM_007585 Anxa2 2.1 5.9 3.1 annexin A2 AV152334 Atp1b1 -3.1 -5.1 -11.9 ATPase, Na+/K+ transporting, beta 1 polypeptide AV306734 B3gnt1 2.8 2.1 2.4 UDP-GlcNAc:betaGal beta-1,3-N-acetylglucosaminyltransferase 1 NM_133348 Bach-pending 4.1 5.9 5.2 brain acyl-CoA hydrolase NM_016767 Batf 2.1 3.2 3 basic leucine zipper transcription factor, ATF-like X17502 Bcat1 2.5 2.9 4.9 branched chain aminotransferase 1, cytosolic L16462 Bcl2a1a 7.7 8.1 11.1 B-cell leukemia/lymphoma 2 related protein A1a NM_007551 Blr1 22.5 3.4 11 Burkitt lymphoma receptor 1 NM_019992 Brdg1-pending 3.2 2.9 2.3 BCR downstream signaling 1 NM_007599 Capg 3.3 10.5 2.2 capping protein (actin filament), gelsolin-like NM_009801 Car2 -2.7 -2.8 -3.7 carbonic anhydrase 2 BC008152 Casp1 3.6 3.5 4.7 caspase 1 BG070529 Casp3 3.3 4.4 5.3 caspase 3, apoptosis related cysteine protease NM_007609 Casp4 6.7 5.4 8.4 caspase 4, apoptosis-related cysteine protease NM_007611 Casp7 2.1 2.6 2.8 caspase 7 AK013312 Ccnb2 3.1 6.9 19.8 cyclin B2 AJ131357 Ccr9 -4.8 -4.1 -5.7 chemokine (C-C motif) receptor 9 NM_009856 Cd83 3.4 16 8.9 CD83 antigen U34882 Cd8b -5.2 -3.4 -4.1 CD8 antigen, beta chain NM_007672 Cdr2 -2.5 -2.8 -2.6 cerebellar degeneration-related 2 NM_016904 Cks1 3 4.1 10.2 CDC28 protein kinase 1 BB779100 Csda 4.3 6.7 9.1 cold shock domain protein A J02583 Ctsl -4.4 -2.6 -8.7 cathepsin L NM_009985 Ctsw 4.6 7 -2.6 cathepsin W NM_009910 Cxcr3 6.3 6.2 4.6 chemokine (C-X-C motif) receptor 3 BM208097 Daf1 -4.4 -5.1 -7.1 decay accelerating factor 1 AK015692 Decr1 4.2 2.9 3.5 2,4-dienoyl CoA reductase 1, mitochondrial BB160593 Dntt -6.6 -6.9 -5.6 deoxynucleotidyltransferase, terminal X06746 Egr2 9.3 6.6 5.2 early growth response 2 BB118974 Flot2 -2.3 -2.3 -2.2 flotillin 2 L23636 Flt3l -2.3 -2.8 -2.3 FMS-like tyrosine kinase 3 ligand NM_022888 Folr4 6.2 8.3 4.6 folate receptor 4 (delta) BC027314 G7e-pending 2.5 4 5.1 G7e protein BG965198 Galnt9 -2.8 -4.3 -3.2 UDP-N-acetyl-alpha-D-galactosamine:polypeptide N-acetylgalactosaminyltransferase 9 NM_011820 Ggtla1 -2.4 -3.9 -3.3 gamma-glutamyltransferase-like activity 1 AF201289 Gilz -7.3 -7.8 -21.5 glucocorticoid-induced leucine zipper NM_008253 Hmgb3 2.8 2.4 4 high mobility group box 3 NM_015818 Hs6st1 -2.2 -2.6 -3.1 heparan sulfate 6-O-sulfotransferase 1 AB023132 Icos 5.6 10.1 6.5 inducible T-cell co-stimulator NM_008368 Il2rb 3.7 9.8 4.5 interleukin 2 receptor, beta chain NM_013674 Irf4 4.3 6.4 6 interferon regulatory factor 4 U13371 Kdt1 7.7 14.3 4.5 kidney cell line derived transcript 1 NM_008479 Lag3 11.8 6.9 2.2 lymphocyte-activation gene 3 BC018559 Litaf 5.6 4.2 2.8 LPS-induced TN factor BB311034 Mapre2 4.2 3.8 4.2 microtubule-associated protein, RP/EB family, member 2 AV110584 Mlp 5.5 5.5 6.2 MARCKS-like protein AF004023 Mox2 3.4 2.4 2.2 antigen identified by monoclonal antibody MRC OX-2 NM_022429 Ms4a4c 17.2 6 5.8 membrane-spanning 4-domains, subfamily A, member 4C NM_008677 Ncf4 2.9 4.9 2.9 neutrophil cytosolic factor 4 AV309418 Ndr1 3.6 7.2 -2.6 N-myc downstream regulated 1 BB759157 Nedd4l -2.4 -2.7 -3.1 neural precursor cell expressed, developmentally down-regulated gene 4-like BB311061 Nelf 5.6 4.1 4.4 nasal embryonic LHRH factor NM_008798 Pdcd1 16.9 9.7 4.9 programmed cell death 1 NM_021396 Pdcd1lg2 3.2 3.4 2.3 programmed cell death 1 ligand 2 BF319989 Plscr1 3.9 3.5 6.8 1 M23182 Prf1 -2.2 -2.3 -3 perforin 1 (pore forming protein) BC005440 Ptger2 9.2 7.4 4.3 prostaglandin E receptor 2 (subtype EP2) NM_019417 Ril-pending -6.9 -35.2 -25.3 reversion induced LIM gene NM_009122 Satb1 -2.1 -3.5 -3.2 special AT-rich sequence binding protein 1 NM_019436 Sit-pending -2.2 -2.2 -2.2 SHP2 interacting transmembrane adaptor NM_026052 Smc6l1 -2.7 -3.4 -2.5 SMC6 structural maintenance of 6-like 1 (yeast) AW552416 Sypl 2.2 2.9 4 synaptophysin-like protein AK010186 Tcte1l 2.7 3 2.7 t-complex-associated-testis-expressed 1-like NM_009384 Tiam1 10.8 17.1 8.4 T-cell lymphoma invasion and metastasis 1 NM_013869 Tnfrsf19 4.8 2.1 3.8 tumor necrosis factor receptor superfamily, member 19 NM_011659 Tnfrsf4 7.2 25.7 6.5 tumor necrosis factor receptor superfamily, member 4 BC003475 Tubb2 -5.9 -4.5 -6.8 tubulin, beta 2 AV303043 Wnt5b -7.1 -8.1 -7.6 wingless-related MMTV integration site 5B BQ174301 7-Sep -2.3 -3.1 -3.2 septin 6 AA500897 Axl 3.1 3.3 AXL BM118679 Ccnd2 2.3 2.8 cyclin D2 NM_007720 Ccr8 2.6 6 chemokine (C-C motif) receptor 8 AU045688 Cd160 4.4 2.6 CD160 antigen BB256012 Cd38 4.1 5.3 CD38 antigen NM_007650 Cd5 4.9 4 CD5 antigen BB148221 Csprs 3 2.9 component of Sp100-rs NM_009977 Cst7 4.5 4.1 cystatin F (leukocystatin) AF301018 Cxcr6 3.8 9 chemokine (C-X-C motif) receptor 6 NM_007933 Eno3 8.4 20.4 enolase 3, beta muscle NM_010145 Ephx1 6.1 4.6 1, microsomal AF297615 Ggta1 3.3 3.9 glycoprotein galactosyltransferase alpha 1, 3 AK016567 Gpm6b 15.4 2.8 glycoprotein m6b BB110067 Gpr83 4.7 14.1 G protein-coupled receptor 83 NM_013542 Gzmb 2.3 23.7 granzyme B U04710 Igf2r 3.2 4.3 insulin-like growth factor 2 receptor NM_010555 Il1r2 4.5 16 interleukin 1 receptor, type II AV298107 Klhdc2 2.5 2.8 kelch domain containing 2 NM_133664 Lad1 2.6 6.4 ladinin BC002070 Ly6a 6.6 21.2 lymphocyte antigen 6 complex, locus A NM_022018 Niban 3.6 11.2 niban protein NM_024253 Nkg7 6.5 16.5 natural killer cell group 7 sequence AK002673 Nrp 6.9 13.3 neuropilin BC018470 Oas1g 3.3 3 2'-5' oligoadenylate synthetase 1G M13227 Penk1 6.7 14.6 preproenkephalin 1 NM_009402 Pglyrp 3.8 9.5 peptidoglycan recognition protein AF147785 Plagl1 3.1 8.3 pleiomorphic adenoma gene-like 1 NM_021400 Prg4 2.3 7.3 proteoglycan 4 (megakaryocyte stimulating factor, articular superficial zone protein) NM_011170 Prnp 3.8 4.7 prion protein NM_023380 Samsn1 2.6 4.4 SAM domain, SH3 domain and nuclear localisation signals, 1 NM_011370 Shyc 5.7 8.8 selective hybridizing clone NM_013667 Slc22a2 4.4 5.8 solute carrier family 22 (organic cation transporter), member 2 NM_019507 Tbx21 2.6 4.1 T-box 21 X03802 Tcrg-V2 3.6 2.7 T-cell receptor gamma, variable 2 NM_011558 Tcrg-V4 7.2 4.2 T-cell receptor gamma, variable 4 AV028402 Thy1 -2.1 -2.1 thymus cell antigen 1, theta AF229434 Tnfrsf18 2.6 7.1 tumor necrosis factor receptor superfamily, member 18 M59378 Tnfrsf1b 2.4 6.7 tumor necrosis factor receptor superfamily, member 1b BC028507 Tnfrsf9 17.3 24.8 tumor necrosis factor receptor superfamily, member 9 BB400695 Twg-pending 2.1 2.5 twisted gastrulation protein BM225280 Zfp52 2.5 2.2 zinc finger protein 52 BC021490 Dapk1 -2.7 -2.9 death associated protein kinase 1 NM_010161 Evi2 2.5 3.7 ecotropic viral integration site 2 AF057287 Hrb 2.6 2.1 HIV-1 Rev binding protein BF019883 Idb2 7 3.3 inhibitor of DNA binding 2 NM_021782 Il21 8.6 4.3 interleukin 21 NM_021283 Il4 3 2.4 interleukin 4 NM_008452 Klf2 -2.3 -4.3 Kruppel-like factor 2 (lung) NM_030710 Ly108 3.8 4.9 lymphocyte antigen 108 AV361868 NEW1CP -2.1 -3 NEW1 domain containing protein NM_011136 Pou2af1 3.8 4.4 POU domain, class 2, associating factor 1 NM_008908 Ppic 6.4 7.5 peptidylprolyl isomerase C NM_019662 Rrad -2.5 -2.5 Ras-related associated with diabetes AF305501 Slc29a1 4 2.2 solute carrier family 29 (nucleoside transporters), member 1 NM_019571 Tm4sf9-pending 2.2 2.3 transmembrane 4 superfamily member 9 NM_011613 Tnfsf11 7.3 9.4 tumor necrosis factor (ligand) superfamily, member 11 NM_019418 Tnfsf14 4.1 4 tumor necrosis factor (ligand) superfamily, member 14 NM_011920 Abcg2 -2.5 ATP-binding cassette, sub-family G (WHITE), member 2 NM_007490 Art2a 3.6 ADP-ribosyltransferase 2a NM_013653 Ccl5 4.6 chemokine (C-C motif) ligand 5 NM_010074 Dpp4 2.5 dipeptidylpeptidase 4 L11330 Dusp2 2.2 dual specificity phosphatase 2 NM_007913 Egr1 4.2 early growth response 1 NM_007974 F2rl1 2.1 coagulation factor II (thrombin) receptor-like 1 K00083 Ifng 4 interferon gamma BM207920 Ldlr -2.1 low density lipoprotein receptor AK018038 Lef1 2.3 lymphoid enhancer binding factor 1 BC021640 Mapk12 2.6 mitogen-activated protein kinase 12 AV215438 Mlp 2.2 MARCKS-like protein NM_008668 Nab2 3.8 Ngfi-A binding protein 2 AK002933 Nrgn 5.7 neurogranin NM_011020 Osp94 2.9 osmotic stress protein NM_019492 Rgs3 2.5 regulator of G-protein signaling 3 BC010198 Sh3bp2 2.1 SH3-domain binding protein 2 AK007893 Sostl 5.4 sclerostin-like NM_011896 Spry1 2.2 sprouty homolog 1 (Drosophila) BB550124 Tgm2 3.3 transglutaminase 2, C polypeptide NM_009452 Tnfsf4 4.5 tumor necrosis factor (ligand) superfamily, member 4 NM_010177 Tnfsf6 3.4 tumor necrosis factor (ligand) superfamily, member 6 NM_019811 Acas2 -3.7 -3.6 acetyl-Coenzyme A synthetase 2 (ADP forming) AK008391 Acp5 -2.7 -3.9 acid phosphatase 5, tartrate resistant BC005490 App -2.8 -2.3 amyloid beta (A4) precursor protein NM_007512 Atpi 2.9 2.8 ATPase inhibitor NM_009745 Bcl7b -2.1 -2.6 B-cell CLL/lymphoma 7B NM_020603 Bing4 -2.1 -2.2 BING4 protein BC004702 Birc5 7.4 20.9 baculoviral IAP repeat-containing 5 AF127766 Capn3 3.8 4.7 calpain 3 NM_009828 Ccna2 4.9 13.4 cyclin A2 X58708 Ccnb1 4.1 24.6 cyclin B1 BB204380 Ccr7 -2.8 -3.4 chemokine (C-C motif) receptor 7 NM_007659 Cdc2a 4.6 13 cell division cycle 2 homolog A (S. pombe) BC027026 Cdkn2c 2.1 2.2 cyclin-dependent kinase inhibitor 2C (p18, inhibits CDK4) C85740 Chek1 2.8 4.7 checkpoint kinase 1 homolog (S. pombe) NM_007704 Cipp -2.4 -3.5 channel-interacting PDZ domain protein NM_019465 Crtam -2.5 -2.6 cytotoxic and regulatory T cell molecule NM_015766 Ebi3 5.8 9.7 Epstein-Barr virus induced gene 3 BB284358 Egln3 2.2 5.3 EGL nine homolog 3 (C. elegans) NM_010119 Ehd1 3.9 2.8 EH-domain containing 1 NM_021891 Fignl1 3.7 8.9 fidgetin-like 1 AK010420 Gadd45b 2.5 2.4 growth arrest and DNA-damage-inducible 45 beta BC003830 H1f0 2.2 2.5 H1 histone family, member 0 AY074806 H2afz 2.3 2.4 H2A histone family, member Z AV277326 Hcapg-pending 4.1 11.1 condensation protein G NM_008234 Hells 3.7 9.8 helicase, lymphoid specific AV297651 Hist3h2bb -3.6 -3.3 histone 3, H2bb AV018952 Hmgn3 4.5 2.7 high mobility group nucleosomal binding domain 3 BC019836 Igfbp4 -11.1 -16.7 insulin-like growth factor binding protein 4 AI642438 Lgals1 4.7 3.9 lectin, galactose binding, soluble 1 NM_016753 Lxn 2.1 2.6 latexin BF226166 Madh7 -2.1 -4 MAD homolog 7 (Drosophila) AA396586 Map17-pending 2.2 2.6 membrane-associated protein 17 BB099487 Mcmd6 5.2 5.7 mini chromosome maintenance deficient 6 (S. cerevisiae) NM_011234 Rad51 3.2 8.6 RAD51 homolog (S. cerevisiae) AK015881 Rog-pending 2.8 2.2 repressor of GATA NM_009103 Rrm1 2.4 3.5 ribonucleotide reductase M1 AV301324 Rrm2 11.7 36.5 ribonucleotide reductase M2 AK002313 Serhl 2.4 2.3 serine hydrolase-like AK004519 Sh3bgrl 2.7 3.4 SH3-binding domain glutamic acid-rich protein like NM_011369 Shcbp1 3.7 14.4 Shc SH2-domain binding protein 1 AW552416 Sypl 3 4.8 synaptophysin-like protein NM_009387 Tk1 4.5 12.5 thymidine kinase 1 BM211413 Top2a 7.3 13.9 topoisomerase (DNA) II alpha BC020139 Tyms 4.3 12.3 thymidylate synthase AV162459 Ube2c 4.1 13.8 ubiquitin-conjugating E2C AV171029 Ubqln2 -2.3 -2.8 ubiquilin 2 BC001990 Unc119h 2.8 3.8 unc119 homolog (C. elegans) NM_019736 Acate2-pending 2.1 acyl-Coenzyme A thioesterase 2, mitochondrial NM_022816 Acate3-pending 2.1 acyl-Coenzyme A thioesterase 3, mitochondrial NM_009610 Actg2 2.4 actin, gamma 2, smooth muscle, enteric NM_007403 Adam8 2.4 a disintegrin and metalloprotease domain 8 BC013477 Adh1 -6.8 alcohol dehydrogenase 1 (class I) NM_013464 Ahr 11.6 aryl-hydrocarbon receptor NM_013471 Anxa4 4.1 annexin A4 BM228788 Bcl2l 2.3 Bcl2-like BC011092 Ccr1 2.4 chemokine (C-C motif) receptor 1 NM_009915 Ccr2 9.6 chemokine (C-C) receptor 2 D83648 Ccr5 2.4 chemokine (C-C motif) receptor 5 NM_009835 Ccr6 2.8 chemokine (C-C motif) receptor 6 AJ318863 Ccrl2 2.4 chemokine (C-C motif) receptor-like 2 NM_023044 Ci1-pending 7.4 cAMP inducible gene 1 NM_009895 Cish 5.8 cytokine inducible SH2-containing protein NM_013498 Crem 2.3 cAMP responsive element modulator NM_009843 Ctla4 16.2 cytotoxic T-lymphocyte-associated protein 4 NM_022325 Ctsz 2.2 NM_021274 Cxcl10 2.2 chemokine (C-X-C motif) ligand 10 BC013766 Dnajc3 -2.3 DnaJ (Hsp40) homolog, subfamily C, member 3 NM_007899 Ecm1 11.9 extracellular matrix protein 1 NM_010330 Emb -3.9 embigin BM120053 Enc1 -4.5 ectodermal-neural cortex 1 BE307478 Entpd1 3.1 ectonucleoside triphosphate diphosphohydrolase 1 AV376291 Entpd5 -2.9 ectonucleoside triphosphate diphosphohydrolase 5 BC002008 Fabp5 11 fatty acid binding protein 5, epidermal BF136544 Fgl2 7 fibrinogen-like protein 2 NM_054039 Foxp3 21.3 forkhead box P3 AK007410 Gadd45g 3.8 growth arrest and DNA-damage-inducible 45 gamma NM_008089 Gata1 2.2 GATA binding protein 1 NM_018734 Gbp3 2.8 guanylate nucleotide binding protein 3 AV166504 Grn 2.9 granulin BC008626 Icam1 2.1 intercellular adhesion molecule BC005450 Icsbp 2.9 interferon concensus sequence binding protein NM_008321 Idb3 4.5 inhibitor of DNA binding 3 NM_008322 Idh2 -2.6 isocitrate dehydrogenase 2 (NADP+), mitochondrial AF329485 IFGP1 3.4 IFGP1 NM_010548 Il10 2.8 interleukin 10 NM_008353 Il12rb1 2.4 interleukin 12 receptor, beta 1 D13695 Il1rl1 8.9 interleukin 1 receptor-like 1 AF054581 Il2ra 18.3 interleukin 2 receptor, alpha chain NM_012057 Irf5 2.1 interferon regulatory factor 5 NM_008399 Itgae 36.9 integrin, alpha E, epithelial-associated X65997 Kit 2.7 kit oncogene NM_016970 Klrg1 23.5 killer cell lectin-like receptor subfamily G, member 1 AU024771 Laptm4b 2.7 lysosomal-associated protein transmembrane 4B NM_011175 Lgmn 3.2 legumain NM_019390 Lmna 3.4 lamin A NM_010728 Lox 4.3 lysyl oxidase NM_015763 Lpin1 -2.6 lipin 1 NM_010735 Lta 3.7 lymphotoxin A NM_008519 Ltb4r1 2.6 leukotriene B4 receptor 1 AF236118 Mona -2.7 monocytic adaptor BQ176602 Myo1c 3.2 myosin IC NM_022995 N4wbp4-pending 2.9 Nedd4 WW binding protein 4 AF155372 Nfkb2 2.7 nuclear factor of kappa light polypeptide gene enhancer in B-cells 2, p49/p100 NM_023893 Ng23-pending 2.3 Ng23-pending NM_008741 Nsg2 -4.4 neuron specific gene family member 2 AV213379 Oxct -2.1 3-oxoacid CoA transferase AF114437 Pctp 4.1 phosphatidylcholine transfer protein NM_138606 Pim2 -2.2 proviral integration site 2 AK012816 Plp2 3 proteolipid protein 2 NM_008897 Pon3 2.2 paraoxonase 3 NM_007548 Prdm1 4.3 PR domain containing 1, with ZNF domain BB274009 Prkar1b 3.6 protein kinase, cAMP dependent regulatory, type I beta NM_011196 Ptger3 2.2 prostaglandin E receptor 3 (subtype EP3) BC011193 Ptger4 3.1 prostaglandin E receptor 4 (subtype EP4) NM_009003 Rab4a -2.2 RAB4A, member RAS oncogene family M34476 Rarg 2.5 retinoic acid receptor, gamma NM_009046 Relb 2.5 avian reticuloendotheliosis viral (v-rel) oncogene related B U94828 Rgs16 2.5 regulator of G-protein signaling 16 BC005569 Rnase4 8.8 ribonuclease, RNase A family 4 D00208 S100a4 27.1 S100 calcium binding protein A4 NM_011313 S100a6 30 S100 calcium binding protein A6 (calcyclin) AB016248 Sc5d 3 sterol-C5-desaturase (fungal ERG3, delta-5-desaturase) homolog (S. cerevisae) NM_011521 Sdc4 2.8 syndecan 4 NM_013658 Sema4a -2.8 sema domain, immunoglobulin domain (Ig), transmembrane domain (TM) and short cytoplasmic domain, (semaphorin) 4A BC002065 Serpina3g 34 serine (or cysteine) proteinase inhibitor, clade A, member 3G NM_011361 Sgk -6.3 serum/glucocorticoid regulated kinase NM_011374 Siat8a -5.5 sialyltransferase 8 (alpha-2, 8-sialyltransferase) A BB414515 Slc2a3 2.2 solute carrier family 2 (facilitated glucose transporter), member 3 BC004831 Smox-pending 2.7 spermine oxidase AK004449 Snx9 2.3 sorting nexin 9 NM_007706 Socs2 16.8 suppressor of cytokine signaling 2 AK019882 Swap70 9.8 SWAP complex protein, 70 kDa BC004829 Syngr2 2.4 synaptogyrin 2 NM_009305 Syp 2.5 synaptophysin BI081723 Tc10-pending 2.6 ras-like protein NM_009332 Tcf3 2.1 transcription factor 3 BM248342 Tgfbr1 4.9 transforming growth factor, beta receptor I BG064103 Traf1 4 Tnf receptor-associated factor 1 BF682223 Ugcg -2.7 UDP-glucose ceramide glucosyltransferase AK009266 Vamp5 2.4 vesicle-associated membrane protein 5 NM_016895 Ak2 2.1 adenylate kinase 2 BC027319 Atp1b1 -4 ATPase, Na+/K+ transporting, beta 1 polypeptide AV149605 Blmh 2.4 bleomycin hydrolase AF002823 Bub1 7.5 budding uninhibited by benzimidazoles 1 homolog (S. cerevisiae) NM_009773 Bub1b 6.3 budding uninhibited by benzimidazoles 1 homolog, beta (S. cerevisiae) X66083 Cd44 3.3 CD44 antigen BB041150 Cdc20 8.3 cell division cycle 20 homolog (S. cerevisiae) NM_009862 Cdc45l 3.5 cell division cycle 45 homolog (S. cerevisiae)-like NM_009870 Cdk4 2.2 cyclin-dependent kinase 4 AV132173 Cenpa 5 centromere autoantigen A BC025084 Cenph 2.4 centromere autoantigen H NM_007691 Chek1 2.9 checkpoint kinase 1 homolog (S. pombe) NM_018763 Chst2 2.2 carbohydrate sulfotransferase 2 BB724741 Cnn3 -8.1 calponin 3, acidic NM_023210 Cpd1-pending 2.5 cerebellar postnatal development protein 1 BC027426 Creg -2.2 cellular repressor of E1A-stimulated genes BG064656 Ctla2b 2.1 cytotoxic T lymphocyte-associated protein 2 beta NM_009983 Ctsd -2.2 cathepsin D AY027937 Ddb2 -2.1 damage specific DNA binding protein 2 AF091101 Dutp 3.9 deoxyuridine triphosphatase NM_007900 Ect2 4.1 ect2 oncogene NM_021398 Eeg1-pending 2.3 embryonic epithelial gene 1 NM_007915 Ei24 2.1 etoposide induced 2.4 mRNA BB251459 Eif4a2 5.2 eukaryotic translation initiation factor 4A2 BC005486 Ets2 -4.7 E26 avian leukemia oncogene 2, 3' domain NM_019739 Foxo1 -3.7 forkhead box O1 BB021390 Foxp1 -3.2 forkhead box P1 AK011088 G1rzfp-pending -3.2 g1-related zinc finger protein NM_010260 Gbp2 -4.5 guanylate nucleotide binding protein 2 AI391218 Glul -2.1 glutamate-ammonia ligase (glutamine synthase) NM_020567 Gmnn 4.4 geminin BB777344 Gnpi 2.5 glucosamine-6-phosphate deaminase BI081061 Grcc8 5.7 gene rich cluster, C8 gene BC012707 Gstt2 4.7 glutathione S-transferase, theta 2 BF319015 Gtpat12 -2.5 gene trap PAT 12 NM_013755 Gyg1 2.9 glycogenin 1 M34962 H2-L -2.1 histocompatibility 2, L region M29881 H2-Q7 -2.3 histocompatibility 2, Q region locus 7 BI690696 Hmgcs1 2.2 3-hydroxy-3-methylglutaryl-Coenzyme A synthase 1 NM_011940 Ifi202b 4.5 interferon activated gene 202B NM_018738 Igtp -2.8 interferon gamma induced GTPase AF065914 Il2 2.2 interleukin 2 AB049137 Il21r -2.3 interleukin 21 receptor NM_010557 Il4ra -7.3 interleukin 4 receptor, alpha NM_008372 Il7r -5.6 interleukin 7 receptor AV319348 Ilk -2.5 integrin linked kinase BM234447 Kif11 4.8 kinesin family member 11 NM_019499 Mad2l1 5.2 MAD2 (mitotic arrest deficient, homolog)-like 1 (yeast) BB100920 Marcks 2.8 myristoylated alanine rich protein kinase C substrate BC011482 Mbc2 -2.3 membrane bound C2 domain containing protein NM_010790 Melk 5.4 maternal embryonic leucine zipper kinase NM_010858 Myla 3.6 myosin light chain, alkali, cardiac atria BE285362 Ncbp2 2.4 nuclear cap binding protein subunit 2 NM_010931 Np95 8 nuclear protein 95 AW908839 Nr3c1 -2.4 nuclear receptor subfamily 3, group C, member 1 NM_008556 Pea15 -2.1 phosphoprotein enriched in astrocytes 15 NM_008828 Pgk1 2.3 phosphoglycerate kinase 1 NM_011121 Plk 4.5 polo-like kinase homolog, (Drosophila) NM_008903 Ppap2a 4 phosphatidic acid phosphatase 2a NM_011563 Prdx2 2.2 peroxiredoxin 2 BC004827 Psat-pending 6 phosphoserine aminotransferase BC027445 Ptp4a3 -2.7 protein tyrosine phosphatase 4a3 AF069051 Pttg1 7.2 pituitary tumor-transforming 1 BB699846 Rab6 -2.1 RAB6, member RAS oncogene family NM_020296 Rbms1 -2.9 RNA binding motif, single stranded interacting protein 1 NM_019955 Ripk3 2.1 receptor-interacting serine-threonine kinase 3 BC019126 Rpe 3 ribulose-5-phosphate-3-epimerase BF119714 Rrm2 11.3 ribonucleotide reductase M2 BC025837 Sbk-pending -2.1 SH3-binding kinase NM_011364 Sh2d1a 2.2 SH2 domain protein 1A BC004585 Sil 2.6 Tal1 interrupting locus NM_009193 Slbp 3 stem-loop binding protein NM_007707 Socs3 -2.8 suppressor of cytokine signaling 3 NM_009270 Sqle 2.3 squalene epoxidase NM_024186 Ssbp2 -3.4 single-stranded DNA binding protein 2 BB246854 Stard4 3.3 StAR-related lipid transfer (START) domain containing 4 BC003261 Stk12 6.5 serine/threonine kinase 12 U80932 Stk6 3.8 serine/threonine kinase 6 BC010581 Stmn1 2.4 stathmin 1 NM_011524 Tacc3 6.6 transforming, acidic coiled-coil containing protein 3 NM_011571 Tesk1 -3 testis specific protein kinase 1 BG075396 Tfdp1 3.4 transcription factor Dp 1 BC008556 Thra -6 thyroid alpha L26349 Tnfrsf1a -2.1 tumor necrosis factor receptor superfamily, member 1a AF329969 Tnfrsf25 -3 tumor necrosis factor receptor superfamily, member 25 NM_023209 Topk-pending 7.1 T-LAK cell-originated protein kinase AK010336 Trip13 10.4 thyroid hormone receptor interactor 13 NM_009445 Ttk 3.4 Ttk protein kinase NM_009518 Wnt10a 3.8 wingless related MMTV integration site 10a