Peripheral Proinflammatory Immunization Regulatory T Cells upon+Adaptive Foxp3
Cutting Edge: Intrathymic Differentiation of
and Jocelyne DemengeotMaria F. Moraes-Fontes, Angelina M. Bilate, Juan J. LafaillePaiva, Andreia C. Lino, Ana C. Martins, João H. Duarte, Santiago Zelenay, Marie-Louise Bergman, Ricardo Sousa
Cutting Edge: Intrathymic Differentiation of AdaptiveFoxp3+ Regulatory T Cells upon PeripheralProinflammatory ImmunizationSantiago Zelenay,*,1 Marie-Louise Bergman,* Ricardo Sousa Paiva,* Andreia C. Lino,*Ana C. Martins,* Joao H. Duarte,* Maria F. Moraes-Fontes,* Angelina M. Bilate,†
Juan J. Lafaille,† and Jocelyne Demengeot*
Thymocytes differentiate into CD4+ Foxp3+ regulatoryT cells (TR) upon interaction between their TCR andpeptide–MHC II complexes locally expressed in thethymus. Conversion of naive CD4+ T cells into TR canadditionally take place in the periphery under nonin-flammatory conditions of Ag encounter. In this study,making use of TCR transgenic models naturally devoidof Foxp3+ cells, we report de novo generation of TR upona single footpad injection of Ag mixed with a classicproinflammatory adjuvant. Abrupt TR differentiationupon immunization occurred intrathymically and wasessential for robust tolerance induction in a mousemodel of spontaneous encephalomyelitis. This phenom-enon could be attributed to a specific feature of thymo-cytes, which, in contrast to mature peripheral CD4+
T cells, were insensitive to the inhibitory effects of IL-6on the induction of Foxp3 expression. Our findings un-cover a pathway for TR generation with major implica-tions for immunity and tolerance induction. TheJournal of Immunology, 2010, 185: 3829–3833.
Thymocytes expressing TCRs specific for self Ags pre-sented in the thymus die through the process of neg-ative selection (1) or differentiate into Foxp3+ regu-
latory T cells (TR) (2). Autoreactive thymocytes escaping theseselection events are exported to peripheral organs and tissues,where TR control their activation and pathogenic potential(2). In addition to thymic production of “natural” TR, pe-ripheral CD4+ T cells can differentiate into “adaptive” Foxp3+
cells in a TGF-b–dependent manner upon Ag encounter inthe absence of inflammation in vivo (3, 4). For example,peripheral TR conversion has been observed following ad-
ministration of free Ag by osmotic pumps (5), by the oralroute (6), or by specifically targeting the Ag to dendritic cellsin the absence of adjuvants (7). Proinflammatory and/or ef-fector Th cell cytokines, such as IL-6, IFN-g, or IL-4, inhibitperipheral TR conversion and promote instead the differen-tiation into IL-17–, IFN-g–, or IL-4–producing effector Thcells (8–10).Negative selection and TR differentiation in the thymus
are believed to be restricted to T cells interacting with Agsexpressed locally, including many tissue-specific proteins ec-topically expressed by medullary thymic epithelial cells (11,12). However, recent studies indicate that abundant blood-borne Ags and peripheral Ag-loaded dendritic cells commonlyreach the thymus (13, 14) and participate not only in negativeselection but also in differentiation of Foxp3+ cells (15). Yet,it is not known whether intrathymic differentiation of TR
occurs following peripheral administration of Ag mixed withstandard adjuvants and whether this process is affected byongoing inflammation. Clarifying these issues is essential todetermine the full dynamic and outcome of immune re-sponses to self and non-self peripheral Ags.In this study, using monoclonal TCR transgenic (Tg) mice
naturally devoid of Foxp3+ cells, we show that Ags mixed withCFA and administered by a single footpad injection can reachthe thymus and locally promote the differentiation of Ag-specific thymocytes into Foxp3+ cells, despite the inflamma-tory conditions. This event could be attributed to a specificfeature of thymocytes, which, in contrast to peripheral matureCD4+ T cells, lack surface IL-6R expression and are thusrefractory to the inhibitory effects of IL-6 on the differenti-ation of Foxp3+ cells. Our findings reveal a previously un-recognized mechanism by which the thymus may establishdominant tolerance to self and non-self Ags during the courseof an immune response.
*Instituto Gulbenkian de Ciencia, Oeiras, Portugal; and †Molecular Pathogenesis Pro-gram, Skirball Institute for Biomolecular Medicine, New York University School ofMedicine, New York, NY 10016
1Current address: Immunobiology Laboratory, Cancer Research UK, London ResearchInstitute, Lincoln’s Inn Fields Laboratories, London, United Kingdom.
Received for publication May 6, 2010. Accepted for publication August 10, 2010.
This work was supported by the Fundacao para a Ciencia e Tecnologia, Portugal, andthe Programa Operacional Ciencia e Inovacao with the coparticipation of the FundoComunitario Europeu and the European Union 7th Framework Programme NaturalImmunomodulators as Novel Immunotherapies for Type 1 Diabetes consortium.
Address correspondence and reprint requests to Dr. Santiago Zelenay, Cancer ResearchUK/London Research Institute, Lincoln’s Inn Fields Laboratories, 44 Lincoln’s Inn
Fields, London WC2A 3PX, United Kingdom. E-mail address: [email protected]
The online version of this article contains supplemental material.
Abbreviations used in this paper: BM, bone marrow; CD4SP cell, single-positive T cell;EAE, experimental autoimmune encephalomyelitis; LN, lymph node; MBP, monoclonalanti-myelin basic protein; pMBP, MBP peptide; PTX, pertussis toxin; Tg, transgenic;TgTg, homozygote anti-MBP TCR transgenic mouse; TR, regulatory T cell; T/R2,monoclonal anti-myelin basic protein TCR-specific transgenic RAG-deficient mouse;TXT, thymectomized; WT, wild-type.
Copyright� 2010 by TheAmerican Association of Immunologists, Inc. 0022-1767/10/$16.00
C57BL/10.PL, C57BL/10.PL-Thy1.1, C57BL/10.PL RAG12/2-MBP-TCRTg (T/R2), C57BL/6, and Foxp3GFP reporter knock-in mice were bred atthe Instituto Gulbenkian de Ciencia Animal Facility. Foxp3scurfy, IL-62/2,IL-6+/2, and wild-type (WT) homozygote anti-MBP TCR Tg C57BL/10.PLmice were bred at the Skirball Institute Central Animal Facility, New YorkUniversity Medical Center, New York, NY. Mouse experimental protocolswere approved by the Institutional Ethical Committee and the PortugueseVeterinary General Division. Foxp3GFP knock-in mice (16) were kindlyprovided by A. Rudensky (University of Washington, Seattle, WA).
For immunizations, mice received 100 ml (50 ml in each footpad) of peptideor protein emulsified in CFA (ready-made CFA; Difco/BD Biosciences, SanJose, CA). When indicated, pertussis toxin (PTX; List Biological Laboratories,Campbell, CA) was administrated i.v. in two doses of 200 ng at a 1-d interval.Experimental autoimmune encephalomyelitis (EAE) was monitored as pre-viously described (17). For adoptive transfer, purified cells suspended in 100ml PBS were injected into the retro-orbital plexus. For mixed bone marrow(BM) chimeras, recipient mice were lethally irradiated (900 rads) and re-constituted the following day with T cell-depleted BM cells.
Cell purification and analysis
Cell suspensions from spleen, blood, thymus, or lymph node (LN; popliteal,inguinal, axillary, mesenteric, and brachial) were incubated with a saturatingamount of Fc-block (anti-CD16/CD32) before staining. Nuclear Foxp3 wasdetected according to the manufacturer’s (eBioscience, San Diego, CA) in-structions. For intracellular cytokine staining, cells were stimulated for 4 hwith PMA (50 ng/ml; Sigma-Aldrich, St. Louis, MO) and ionomycin (500ng/ml; Calbiochem, San Diego, CA). Brefeldin A (10 mg/ml; Sigma-Aldrich)was added for the last 2 h of stimulation. Data were acquired on a FACS-Calibur or Aria (BD Biosciences) and analyzed inside a lymphocyte gate withCellQuest (BD Biosciences) and Flowjo (Tree Star, Ashland, OR) software.Cell purification was performed using Aria or MoFlo high-speed cell sorters(DakoCytomation, Carpinteria, CA).
Cell culture
For Foxp3 induction assays, cells were plated at 2.5 3 104 cells/well in flat-bottom 96-well plates with 3 mg/ml plate-bound anti-CD3 mAb, 1 mg/mlsoluble anti-CD28 mAb (eBioscience), ∼10 U/ml IL-2 (X63-IL-2 superna-tant), 0.2 ng/ml TGF-b1, and 20 ng/ml IL-6 (R&D Systems, Minneapolis,MN). Cultures were set in triplicates in a final volume of 200 ml for 72 h. ForStat3 phosphorylation assays, cells were stimulated for 15 min with 100 ng/mlIL-6 and stained with anti-phospho Stat3-AlexaFluor 647 Ab (4/P-Stat3; BDBiosciences), according to the manufacturer’s instructions.
Statistical analysis
Statistical significance was determined using the two-tailed Student t test andthe log rank test. A p value ,0.05 was considered significant (*p , 0.05; **p, 0.01; ***p , 0.001).
Results and DiscussionImmunization promotes tolerance induction through differentiation ofTR
Anti-myelin basic protein (MBP) TCR-specific Tg RAG-deficient mice (T/R2) or homozygotes for the Tg-TCR(TgTg) spontaneously develop severe progressive EAE by 2mo of age and succumb in their third month of life (17). Asthese mice are naturally devoid of Foxp3+ cells, they offer anideal system to test protocols that may induce the de novogeneration of TR. Administration of the nominal Ag togetherwith IFA has been shown to prevent EAE occurrence in T/R2
mice, presumably through the induction of T cell anergy (18).We first tested whether immunization protocols known toinduce various degrees of inflammation would differentiallyaffect disease onset and progression in T/R2 mice. A singlefootpad injection of 100 mg of the agonist N-terminal Ac1-17
MBP peptide (pMBP) mixed with IFA or CFA protectedT/R2 for at least 3 mo, whereas administration of CFA alonedid not alter the course of the disease (Fig. 1A). Protectionfrom EAE was associated with the emergence of Foxp3+ cells,readily detectable in PBLs as early as 6 d postinjection andundetectable in control animals (Fig. 1B, 1C). Foxp3+ cellsreached ∼14% of CD4+ T cells by 6 wk postimmunizationand remained at $4% for an additional 6–7 wk. Full EAEprotection lasted 3 mo, after which immunized mice devel-oped a chronic and mild disease that coincided with a signif-icant decline in peripheral Foxp3+ cell frequency. Renewedinjection of CFA-pMBP 3 mo after the first immunizationprolonged protection to more than 7 mo of age. As CFA isa more potent adjuvant than IFA, it was selected to furtheranalyze the function and origin of Foxp3+ cells generatedupon proinflammatory immunization.Foxp3+ cells emerging upon CFA-pMBP immunization
were phenotypically and functionally bona fide TR. About60% of them expressed CD25, and most CD25+ cells wereFoxp3+, GITR+, and CD103+. They efficiently suppressed theproliferation of conventional CD4+ CD252 cells in vitro anddisplayed regulatory activity in vivo, as shown by adoptivetransfer experiments (Supplemental Fig. 1). Finally, immu-nization did not prevent or delay severe EAE in TgTg bearinga Foxp3 null mutation (Foxp3scurfy) (Fig. 1A), formallydemonstrating that dominant tolerance through de novo in-duction of Foxp3+ TR, and not T cell anergy (18), is themechanism of disease prevention following immunization ofT/R2 mice.
A
B
D
C
FIGURE 1. CFA-peptide immunization promotes tolerance mediated by de
novo generated Foxp3+ TR. A, For each protocol, treatment of mice was
initiated at 1 mo of age and mice were scored for EAE at least twice per week,
until 3 mo of age. Data represent pooled results of two to eight independent
experiments. B and C, T/R2 mice 1 mo old were immunized with CFA-PBS,
CFA-pMBP (B) or with IFA-pMBP (C). Percentage of Foxp3+ cells in
a CD4+ cell gate in PBLs (mean 6 SD; n = 3 per group; d, s) and cu-
mulative EAE incidence (n, N) along time. D, TgTg IL-6+/2 or IL-62/2 mice
1 mo-old were immunized with CFA-pMBP into the footpad and their
draining LNs analyzed 4 or 14 d later. Representative histogram for Foxp3
and percentage of Foxp3+cells measured in a CD4+ cell gate. n = 2–6 in each
group. Data are representative of two independent experiments.
3830 CUTTING EDGE: INTRATHYMIC TR GENERATION UPON IMMUNIZATION
Concomitant differentiation of effector Th cells and TR uponproinflammatory immunization
The above observations needed to be reconciled with the no-tion that CFA is an adjuvant used to induce EAE in WTmice. Our immunization regimen promoted T cell expan-sion, activation, and differentiation into IFN-g– and IL-17–producing cells (Supplemental Fig. 2), all readily detectableby day 4 postinjection and concomitant with Foxp3+ cellemergence. Protocols for efficient induction of EAE in WTanimals commonly rely on the co-administration of PTXand the immunogen. Consistently, administration of PTXabrogated the protective effect of CFA-pMBP (Fig. 1A), am-plified cellular expansion, increased the number of IFN-g–producing cells, and reduced both Foxp3+ cell frequency andnumber (Supplemental Fig. 2). We conclude that whereasCFA-pMBP promotes Th1, Th17, and TR differentiation,PTX inhibits the induction, migration, and/or expansion ofFoxp3+ cells, a role reminiscent of its effect on TR survival andfunction (19).As expected, CFA administration also provoked rapid and
vigorous production of innate cytokines, including IL-6 (notshown). Intriguingly, this cytokine has been shown to playa key role in preventing peripheral TR conversion, notablyupon CFA administration (8). Consistent with this notion,immunized IL-62/2 TgTg mice displayed a 3- to 5-fold in-creased frequency of peripheral Foxp3+ cells when comparedwith control IL-6+/2 TgTg mice similarly treated (Fig. 1D).As unimmunized IL-62/2 and IL-6+/2 TgTg mice weredevoid of Foxp3+ cells, we conclude, as previously shown(8), that inflammation driven by IL-6 interfered with TR
conversion. Collectively, our results suggest that a subset ofT cells in IL-6–competent mice is insensitive to the inhibitoryeffect of IL-6 and consequently can convert to TR despiteongoing inflammation.
We next examined whether Foxp3+ cell differentiation in im-munized T/R2 mice occurred in the thymus. Remarkably, ki-netic analysis in T/R2 mice immunized with a single footpadinjection of CFA-pMBP revealed that Foxp3+ cells represen-ted ∼12% of CD4+ single-positive T (CD4SP) cells in thethymus by day 3 postimmunization (Fig. 2A–C). Noteworthy,Foxp3+ cells were not detectable in LNs before day 4, excludingthe possibility that thymic Foxp3+ cells represented circulatingperipherally differentiated TR (Fig. 2A). Despite an increasingfrequency of thymic TR during 7 d postimmunization, the totalnumber of Foxp3+ thymocytes was highest at day 3 and grad-ually decreased following the reduction in total CD4SP. Incontrast, peripheral Foxp3+ cell frequency increased at leastuntil day 14 postimmunization (Fig. 2B, 2C). Intrathymic dif-ferentiation of TR following footpad immunization was Agspecific and not a singularity of anti-MBP Tg mice, or of smallsynthetic self-peptides, as similar results were obtained uponCFA-peptide or CFA-protein, but not CFA-PBS, administra-tion to anti-OVA DO11.10 RAG12/2 TCR-Tg mice (Sup-plemental Figs. 3, 4). Together, these results demonstrate thatnot only blood-borne (15) but also peripheral Ags administereds.c. in a stable water-in-oil emulsion can enter the thymus andbe presented locally to promote both deletion and TR differ-entiation.
A
B
C
D
E F
FIGURE 2. Peripheral proinflammatory immunization leads to abrupt Ag-
specific TR differentiation in the thymus. A–C, T/R2 mice 1 mo old were
immunized with CFA-pMBP into the footpad and thymus, and LNs were
analyzed at different time points. A, Representative dot plot for Foxp3 versus
CD25 inside a CD4+ CD82 gate from the thymus (upper panel) or from LNs
(lower panel). B, Number of total thymocytes (left) and CD4SP thymocytes
(right). C, Percentage of Foxp3+ cells among CD4SP thymocytes or CD4+
LN cells (left) and number of Foxp3+ CD4SP thymocytes or Foxp3+ CD4+
LN cells (right). B and C are pooled kinetic analysis from five experiments
with 2–12 mice in each time point. D, Lethally irradiated C57BL/10.PL-
Thy1.1 mice were reconstituted with a 9:1 mixture of BM cells from Thy1.2+
T/R2 mice and Thy1.1+ WT mice. Chimeric mice were untreated (n = 3) or
immunized into the footpad with CFA-pMBP (n = 3) 2 mo after re-
constitution and analyzed 4 d later. Number of CD4SP thymocytes (left) andpercentage of Foxp3+ cells among CD4SP cells (right). Data are representative
of three independent experiments; Student t test. **p , 0.01. E and F, One
group of 1-mo-old euthymic T/R2 mice was left untreated (;; n = 12),
although both TXT (s; n = 7) and sham-TXT (d; n = 5) were immunized
with CFA-pMBP 3 d postsurgery. E, Percentage of Foxp3+ cells in a CD4+
cell gate from PBLs over time; Student t test. F, Percentage of survival overtime (p , 0.01, untreated versus TXT CFA-pMBP; p , 0.01, TXT CFA-
We next ascertained that Ag-specific TR differentiate in thethymus following immunization of mice bearing a polyclonalrepertoire of lymphocytes. BM from T/R2 (Thy1.2+) andWT (Thy1.1+) mice were co-injected into lethally irradiatedWT (Thy1.1+) animals. These mixed BM chimeras were im-munized with CFA-pMBP at 2 mo postreconstitution whenthey contained 5–10% of Tg CD4SP cells. Analysis of theirthymi 4 d postimmunization revealed both negative selectionand TR differentiation of Ag-specific CD4SP thymocytes, asindicated by a 100-fold reduction in number and the emer-gence of ∼7% Foxp3+ cells among Thy1.2+ cells. Importantly,among the WT polyclonal Thy1.1+ cells, neither thymocytenumber nor Foxp3+ cell frequency was affected by immuniza-tion (Fig. 2D). These results demonstrate that intrathymic TR
differentiation upon peripheral proinflammatory immuniza-tion is restricted to Ag-specific T cells and can occur in thecontext of a normal polyclonal repertoire of lymphocytes.To directly evaluate the contribution of the thymus to tol-
erance induction and to the accumulation of peripheral TR
upon immunization, healthy 4-wk-old T/R2 mice were thy-mectomized (TXT) and immunized with CFA-pMBP 3 d later.Strikingly, Foxp3+ cells were undetectable in peripheral bloodof TXT mice for the first 2 wk postimmunization. In addi-tion, TR frequency was significantly lower in TXT than ineuthymic mice at any time point of the 10-wk-long kinetic(Fig. 2E). These results are in agreement with our kineticanalysis above, indicating that early and efficient differentia-tion of TR is restricted to the thymus. Nevertheless, immu-nization delayed disease progression in TXT mice, possibly asa result of peripheral generation of functional Foxp3+ cells.However, although all TXT T/R2 mice succumbed to EAEby 30 wk after immunization, at this same time, euthymicanimals were alive and only 40% of them showed signs ofmild disease (score #2) (Fig. 2F). Together, our data dem-onstrate that the thymus, most likely owing to intrathymic TR
differentiation, is essential for induction of potent and long-lasting tolerance.
Immature CD4+ SP thymocytes are refractory to IL-6–mediatedinhibition of Foxp3 induction
Taken together, the above results indicated that intrathymic, incontrast to peripheral, TR differentiation is insensitive to theinhibitory effects of inflammation. We next tested whetherthese differences could be attributed to intrinsic features ofdeveloping T cells by performing in vitro TR differentiationassays. As expected, addition of IL-6 to T/R2 peripheral CD4+
T cells inhibited by ∼70% the generation of Foxp3+ cellsinduced by TGF-b (Fig. 3A). However, this inhibition wassignificantly lower (∼40%) for CD4SP thymocytes. Simi-lar results were obtained when testing WT CD4+ CD82
Foxp32 cells isolated from Foxp3GFP knock-in mice (Fig. 3B,3C). Strikingly, purified HSAhigh Foxp32CD4SP thymocytes,purged of recirculating and more mature cells (see Ref. 16),were totally refractory to the inhibitory effect of IL-6. Thesensitivity of each cell subset to IL-6 directly correlated with thelevel of both surface IL-6Ra expression (Fig. 3D) and proximalIL-6R signaling, determined by intracellular phosphorylatedStat3 detection upon IL-6 exposure (Fig. 3E). Together, theseresults indicate that newly formed T cells are insensitive to IL-6–mediated inhibition of Foxp3 induction owing to theirvery limited responsiveness to this cytokine. These findings
are in agreement with the notion that maturation stage is onekey factor controlling the predisposition of T cells toward TR
differentiation (20) and provide a molecular basis for our ob-servation that de novo generation of TR takes place in thethymus rather than in the periphery upon proinflammatoryimmunization.In conclusion, we show that a single footpad injection of Ags
mixed with a highly inflammatory adjuvant promotes the denovo generation of protective Ag-specific TR. This unexpectedfinding is explained by the demonstration that vigorous TR
differentiation occurred intrathymically early following im-munization. The latter phenomenon occurred despite theelevated levels of inflammatory mediators in circulation, no-tably IL-6, most likely owing to the specific resistance ofnewly formed T cells to the inhibitory effects of IL-6 on TR
differentiation.An essential role of the thymus in ensuring induced pe-
ripheral tolerance has been occasionally reported, and mostlyattributed to a property of recent thymic emigrants (e.g., seeRef. 21). In view of our findings, it is tempting to speculatethat immature T cells exported to the periphery may bethe preferential precursors of peripherally differentiated TR. Arecent study reported increased TR frequency in the thymus of
A B
C
D E
FIGURE 3. Immature CD4SP thymocytes are refractory to IL-6–mediated
inhibition of Foxp3 induction. A, CD4+ CD82 thymocytes or CD4+ LN cells
were isolated from T/R2 mice and cultured with plate-bound anti-CD3, anti-
CD28, TGF-b 6 IL-6. Percentage of Foxp3+ cells with TGF-b + IL-6 rel-
ative to control with only TGF-b. ***p , 0.001. B and C, CD4+ GFP2 LN
cells and either total CD4+ CD82 GFP2 or HSAhigh thymocytes were iso-
lated from Foxp3GFP mice and cultured as in A plus IL-2. B, Representativehistogram of GFP expression in CD4+ T cells. C, Percentage of GFP+ cells
with TGF-b + IL-6 relative to control with only TGF-b. Data are repre-
sentative of three independent experiments. ***p , 0.001. D, Representativehistograms of surface IL-6Ra expression in gated CD4+ GFP2 cells from
non-Tg mice upon induction of EAE by immunization, aphenomenon associated with elevated intrathymic IL-7 ex-pression (22). These thymic TR allowed disease remission,suggesting that they were immunogen specific. Together, thiswork and ours indicate that intrathymic TR differentiationupon peripheral immunization is a robust event, because ofboth microenvironment modifications and, as we demonstratein this study, intrinsic properties of thymocytes.Tolerance induction to peripheral Ags that gain access to the
thymus may be essential not only to purge the repertoire of self-reactive T cells specific to Ags that are not expressed by thymicAPCs (13) but also to broaden the repertoire of Foxp3+ TR,assuring robust dominant tolerance to peripheral tissues. In ad-dition, our evidence that intrathymic differentiation of TR cantake place under strong and systemic inflammatory conditionsuncover a potential mechanism for the emergence of pathogen-specific TR during infections, provided pathogen-derived Agsgain access to the thymus. Noteworthy is that microbe-specificTR have been shown to be essential for protection against a sec-ondary challenge to the microbe (23). Thus, specific TR differ-entiation uponAg access to the thymusmay play an essential rolenot only in the establishment of self-tolerance but also in im-munity to reinfection. Finally, our observations have majorimplications for the design of vaccines for preventing or treatinginfectious diseases, cancer, or autoimmunity.
AcknowledgmentsWe thank R.M. Santos for Ab preparation, A. Perez and R. Gardner for op-
erating the cell sorter, M. Rebelo for mouse colony management, A. Coutinho
and T.L. Carvalho for constructive discussions during the development of this
work, and C. Reis e Sousa for critical reading of the manuscript.
DisclosuresThe authors have no financial conflicts of interest.
References1. Palmer, E. 2003. Negative selection—clearing out the bad apples from the
T-cell repertoire. Nat. Rev. Immunol. 3: 383–391.2. Sakaguchi, S., M. Ono, R. Setoguchi, H. Yagi, S. Hori, Z. Fehervari, J. Shimizu,
T. Takahashi, and T. Nomura. 2006. Foxp3+ CD25+ CD4+ natural regulatoryT cells in dominant self-tolerance and autoimmune disease. Immunol. Rev. 212: 8–27.
3. Kretschmer, K., I. Apostolou, E. Jaeckel, K. Khazaie, and H. von Boehmer. 2006.Making regulatory T cells with defined antigen specificity: role in autoimmunityand cancer. Immunol. Rev. 212: 163–169.
4. Lohr, J., B. Knoechel, and A. K. Abbas. 2006. Regulatory T cells in the periphery.Immunol. Rev. 212: 149–162.
5. Apostolou, I., and H. von Boehmer. 2004. In vivo instruction of suppressorcommitment in naive T cells. J. Exp. Med. 199: 1401–1408.
6. Mucida, D., N. Kutchukhidze, A. Erazo, M. Russo, J. J. Lafaille, and M. A. Curottode Lafaille. 2005. Oral tolerance in the absence of naturally occurring Tregs. J. Clin.Invest. 115: 1923–1933.
7. Kretschmer, K., I. Apostolou, D. Hawiger, K. Khazaie, M. C. Nussenzweig, andH. von Boehmer. 2005. Inducing and expanding regulatory T cell populations byforeign antigen. Nat. Immunol. 6: 1219–1227.
8. Korn, T., M. Mitsdoerffer, A. L. Croxford, A. Awasthi, V. A. Dardalhon, G. Galileos,P. Vollmar, G. L. Stritesky, M. H. Kaplan, A. Waisman, et al. 2008. IL-6 controlsTh17 immunity in vivo by inhibiting the conversion of conventional T cells intoFoxp3+ regulatory T cells. Proc. Natl. Acad. Sci. USA 105: 18460–18465.
9. Wei, J., O. Duramad, O. A. Perng, S. L. Reiner, Y. J. Liu, and F. X. Qin. 2007.Antagonistic nature of T helper 1/2 developmental programs in opposing peripheralinduction of Foxp3+ regulatory T cells. Proc. Natl. Acad. Sci. USA 104: 18169–18174.
10. Caretto, D., S. D. Katzman, A. V. Villarino, E. Gallo, and A. K. Abbas. 2010.Cutting edge: the Th1 response inhibits the generation of peripheral regulatoryT cells. J. Immunol. 184: 30–34.
11. Derbinski, J., A. Schulte, B. Kyewski, and L. Klein. 2001. Promiscuous gene ex-pression in medullary thymic epithelial cells mirrors the peripheral self. Nat.Immunol. 2: 1032–1039.
12. Aschenbrenner, K., L. M. D’Cruz, E. H. Vollmann, M. Hinterberger, J. Emmerich,L. K. Swee, A. Rolink, and L. Klein. 2007. Selection of Foxp3+ regulatory T cellsspecific for self antigen expressed and presented by Aire+ medullary thymicepithelial cells. Nat. Immunol. 8: 351–358.
13. Bonasio, R., M. L. Scimone, P. Schaerli, N. Grabie, A. H. Lichtman, and U. H. vonAndrian. 2006. Clonal deletion of thymocytes by circulating dendritic cells homingto the thymus. Nat. Immunol. 7: 1092–1100.
14. Li, J., J. Park, D. Foss, and I. Goldschneider. 2009. Thymus-homing peripheraldendritic cells constitute two of the three major subsets of dendritic cells in thesteady-state thymus. J. Exp. Med. 206: 607–622.
15. Atibalentja, D. F., C. A. Byersdorfer, and E. R. Unanue. 2009. Thymus-bloodprotein interactions are highly effective in negative selection and regulatory T cellinduction. J. Immunol. 183: 7909–7918.
16. Fontenot, J. D., J. P. Rasmussen, L. M. Williams, J. L. Dooley, A. G. Farr, andA. Y. Rudensky. 2005. Regulatory T cell lineage specification by the forkheadtranscription factor foxp3. Immunity 22: 329–341.
17. Lafaille, J. J., K. Nagashima, M. Katsuki, and S. Tonegawa. 1994. High incidenceof spontaneous autoimmune encephalomyelitis in immunodeficient anti-myelinbasic protein T cell receptor transgenic mice. Cell 78: 399–408.
18. Marusic, S., and S. Tonegawa. 1997. Tolerance induction and autoimmune en-cephalomyelitis amelioration after administration of myelin basic protein-derivedpeptide. J. Exp. Med. 186: 507–515.
19. Chen, X., R. T. Winkler-Pickett, N. H. Carbonetti, J. R. Ortaldo, J. J. Oppenheim,and O. M. Howard. 2006. Pertussis toxin as an adjuvant suppresses the number andfunction of CD4+CD25+ T regulatory cells. Eur. J. Immunol. 36: 671–680.
20. Wirnsberger, G., F. Mair, and L. Klein. 2009. Regulatory T cell differentiation ofthymocytes does not require a dedicated antigen-presenting cell but is under T cell-intrinsic developmental control. Proc. Natl. Acad. Sci. USA 106: 10278–10283.
21. Song, F., Z. Guan, I. E. Gienapp, T. Shawler, J. Benson, and C. C. Whitacre. 2006.The thymus plays a role in oral tolerance in experimental autoimmune encepha-lomyelitis. J. Immunol. 177: 1500–1509.
22. Chen, X., L. Fang, S. Song, T. B. Guo, A. Liu, and J. Z. Zhang. 2009. Thymicregulation of autoimmune disease by accelerated differentiation of Foxp3+ regula-tory T cells through IL-7 signaling pathway. J. Immunol. 183: 6135–6144.
23. Belkaid, Y., C. A. Piccirillo, S. Mendez, E. M. Shevach, and D. L. Sacks. 2002.CD4+CD25+ regulatory T cells control Leishmania major persistence and immu-nity. Nature 420: 502–507.
