A RTICLES

Aire regulates negative selection of

organ-specific T cells
Adrian Liston1, Sylvie Lesage1, Judith Wilson1, Leena Peltonen2 and Christopher C. Goodnow1
© 2003 Nature Publishing Group http://www.nature.com/natureimmunology

Published online 3 March 2003; doi:10.1038/ni906

Autoimmune polyendocrinopathy syndrome type 1 is a recessive Mendelian disorder resulting from

mutations in a novel gene, AIRE, and is characterized by a spectrum of organ-specific autoimmune
diseases. It is not known what tolerance mechanisms are defective as a result of AIRE mutation. By
tracing the fate of autoreactive CD4 + T cells with high affinity for a pancreatic antigen in transgenic
mice with an Aire mutation, we show here that Aire deficiency causes almost complete failure to
delete the organ-specific cells in the thymus.These results indicate that autoimmune polyendocrino-
pathy syndrome 1 is caused by failure of a specialized mechanism for deleting forbidden T cell clones,
establishing a central role for this tolerance mechanism.

Immunological tolerance to organ-specific antigens poses a particular
problem. These antigens are often present in high concentrations in
peripheral organs, where they potentially activate both high- and low-
affinity T cells, but only trace amounts occur in the circulation,
peripheral lymphoid tissues or thymus, where they are needed to
a b c d
induce tolerance. Organ-specific T cells in several different lines of T
cell receptor (TCR) transgenic mice have a complete lack of central
or peripheral tolerance1–6. For these reasons, it is thought that organ-
specific tolerance depends primarily on a passive process of immuno-
logical ‘ignorance’ due to sequestration of the antigen from naive T
cells1, or on peripheral tolerance mechanisms such as peripheral dele-
tion7–9, anergy9, suppression by regulatory T cells10,11 or other, less
well-defined, regulatory mechanisms4,5. A growing body of findings,
however, indicates that the thymus may have an impact on organ-spe-
cific tolerance. When high-affinity T cells recognizing dominant pep-
tide epitopes of model antigens were tracked in TCR transgenic mice
expressing these antigens under organ-specific promoters, the pre-
vailing outcomes are thymic clonal deletion and anergy of the autore-
active T cells12–15. Analysis of normal and transgenic mice showed that
a wide range of ‘organ-specific antigens’ are actually expressed by
rare peripheral antigen expressing (PAE) thymic medullary epithelial
cells16,17. Thus, there are widely divergent views about normal resis-
tance to organ-specific autoimmune diseases.
Although transgenic animal models have been crucial in establish-
ing the possible fates of autoreactive T cells, resolving the divergent
mechanisms of tolerance requires ways to test their significance for
immune regulation in humans and animals with normal lymphocyte
repertoires. Analysis of genetic defects that result in spontaneous
autoimmune disease in humans provides, in principal, a powerful
bridge between human immunoregulation and the transgenic mouse

Figure 1. Failure to delete pancreatic islet reactive T cells in thymus of

Aire–/– mice. T cells bearing the insHEL-reactive TCR were stained with 1G12 anti-
TCR clonotypic antibody and T cell maturation markers and analyzed by four-color
flow cytometry. Representative profiles from mice of the indicated genotypes are
shown. Numbers indicate the percentages of gated cells that fall within the indicated
regions. Histogram profiles show the control B10.Br Aire+/+ TCR transgenic mice (in
white) superimposed on the indicated profile (in gray) to facilitate comparison. (a)
Control B10.Br Aire+/+ TCR transgenic mice, lacking the insHEL gene. (b) Control
B10.Br Aire+/+ TCR-insHEL double-transgenic mice. (c) Aire+/– TCR-insHEL double-
transgenic mice. (d) Aire–/– TCR-insHEL double-transgenic mice.

1
ACRF Genetics Lab, Medical Genome Centre, John Curtin School of Medical Research,The Australian National University, ACT 2601 Australia. 2Department of Medical
Genetics and Molecular Medicine, University of Helsinki and National Public Health Institute, Finland. Correspondence should be addressed to C.C.G.
(chris.goodnow@anu.edu.au).

350 nature immunology • volume 4 no 4 • april 2003 • www.nature.com/natureimmunology

 A RTICLES

Table 1. Loss of Aire rescues autoreactive CD4 T cells from negative selection
Aire genotype B10.Br B10.Br Aire+/+ Aire+/– Aire–/–
Transgene genotype TCR+ TCR+HEL+ TCR+HEL+ TCR+HEL+ TCR+HEL+

CD4+CD8+double positive cellsa 680 ± 130 270 ± 100 270 ± 140 500 ± 230 680 ± 130
CD4+CD8– single positive cellsa 360 ± 60 50 ± 20 44 ± 20 98 ± 40 230 ± 80
CD4+CD8– 1G12hi cellsa † 220 ± 86 9.0 ± 3 4.8 ± 4 39 ± 21 120 ± 42
© 2003 Nature Publishing Group http://www.nature.com/natureimmunology

CD4+CD8– 1G12hi CD69– cellsa 110 ± 51 2.8 ± 1 1.3 ± 1 15 ± 10 61 ± 34

CD4+CD8– 1G12hi CCL19-Rhi cellsa 110 ± 43 1.9 ± 2 0.9 ± 0.8 11 ± 10 59 ± 34
CD4+CD8– 1G12+CD25hi cellsa 7.7 ± 4 8.2 ± 3 4.1 ± 3 7.3 ± 2 3.8 ± 0.7
CD25hi within CD4+CD8– 1G12+ cells (%) 2.3 ± 0.8 24.5 ± 3 16 ± 4 9.7 ± 0.4 2.2 ± 0.7
Mean CD5 on CD4+CD8– 1G12+ cells (relative to TCR+) 100 152 ± 33 129 ± 5 133 ± 12 109 ± 9
Mean CD3 on CD4+CD8– 1G12+ cells (relative to TCR+) 100 50 ± 12 51 ± 9 70 ± 15 86 ± 13
1G12+CD4+ splenocytes (%) 6.1 ± 2.2 0.4 ± 0.1 0.4 ± 0.1 0.6 ± 0.3 2.2 ± 2

Mean ± s.d. aCell no. × 105.

models needed to analyze tolerance mechanisms. A key opportunity is
the genetic defect responsible for human autoimmune polyen-
docrinopathy syndrome type 1, a rare recessive mendelian disorder
characterized by a spectrum of autoimmune diseases involving multi-
ple peripheral organs (reviewed at http://www3.ncbi.nlm.nih.gov,
OMIM 240300). Common conditions include various autoimmune
endocrinopathies, gastrointestinal disorders, skin diseases and chronic
hepatitis, with autoantibodies produced to many organ-specific
autoantigens18. The disease results from loss-of-function mutations in
the gene AIRE19,20. The AIRE protein seems to act as a novel tran-
scriptional regulator21,22 and is chiefly expressed in rare thymus stro-
mal cells23–25. An apt model of human autoimmune polyendocrinopa-
thy syndrome is the Aire-mutant mouse, which shows a range of
autoimmune disorders similar to that observed in humans23. Analysis
of Aire-deficient mice showed that Aire is required for low-level
thymic epithelial expression of a subset of organ-specific mRNAs,
although only a minority of the mRNAs expressed ectopically in
thymic medullary epithelial cells seemed to be dependent on Aire26.
Resolving the role of Aire in organ-specific tolerance awaits a deter-
mination of what, if any, T cell tolerance mechanisms are disrupted by
the gene defect, particularly because thymic medullary expression of
one of the best-characterized examples of a peripheral antigen that
triggers thymic deletion, the liver C-reactive protein15, is not depen-
dent on Aire26.
To understand how Aire mutations affect the acquisition of immuno-
logical tolerance to organ-specific antigens, we intercrossed a targeted
Aire null mutation23 into a TCR transgenic strain that enables direct trac-
ing of organ-specific CD4 T cells14,27,28. The 3A9 TCR transgenic strain
carries rearranged TCRα and TCRβ genes, causing most developing
CD4+ helper T cells to express a single TCR with characterized high
avidity and specificity for a dominant peptide of hen egg lysozyme (HEL)
complexed with the major histocompatibility molecule (MHC) I-Ak (ref.
28). HEL induces immunological tolerance when expressed as a neo-self
antigen from transgenes controlled by systemic or organ-specific promot-
ers29,30. When the membrane HEL transgene is controlled by the rat insulin
promoter (insHEL), high concentrations of antigen are expressed in the β
cells of the pancreatic islet and low concentrations (comparable to those
of insulin) circulate systemically14. In TCR-insHEL double-transgenic
mice, 3A9 T cells are efficiently tolerized despite their high frequency,
through deletion of high-avidity T cells in the thymus and export of resid-
ual cells that are functionally anergic and express lower amounts of HEL-
specific TCR and CD3 and increased amounts of CD514,27. We show here
that Aire mutation causes almost complete failure to delete pancreatic
islet–reactive T cells in the thymus, directly linking autoimmune polyen-
docrinopathy syndrome 1 to failure of thymic deletion and establishing a
central role for this mechanism in immune regulation.
Results
Aire and central tolerance
To determine whether or not Aire has a role in T cell tolerance mecha-
nisms, Aire–/– mice were crossed to 3A9 TCR-insHEL transgenic ani-
mals, enabling direct and sensitive tracking of islet-specific T cells.
Formation of HEL-specific 3A9 T cells was traced by flow cytometric

Figure 2. Failure of thymic deletion in Aire–/– mice. Absolute numbers of

mature CD4 T cells in each thymus were calculated using two independent mea-
sures. Open bars indicate the mean numbers of CD4+CD8– 1G12hiCD69– T cells.
Filled bars indicate the mean numbers of CD4+CD8– 1G12hiCCL19 receptorhi T
cells. Error bars indicate standard deviation. Differences in mature cell numbers
were considered significant if t-test P values (two-sample assuming equal vari-
ances) were below 0.05, for both the number of CD4+CD8–1G12hiCD69– cells and
the number of CD4+CD8–1G12hiCCL19-Rhi cells. By this criterion, B10.Br TCR
mice showed no significant difference from Aire–/– TCR-insHEL mice, and both of
these groups showed substantially more mature cells than all other groups. Aire+/–
TCR-insHEL mice showed significantly higher levels of mature cells than B10.Br
TCR-insHEL mice, but not Aire+/+ TCR-insHEL mice. No other significant differ-
ences were observed.

www.nature.com/natureimmunology • april 2003 • volume 4 no 4 • nature immunology 351

 A RTICLES

a b
© 2003 Nature Publishing Group http://www.nature.com/natureimmunology

Figure 3. Islet-specific T cells in the peripheral repertoire of Aire–/– mice.

Splenocytes were stained with 1G12 clonotypic antibody to TCR and CD4. (a) Repre-
sentative profiles from mice of the indicated genotypes. Numbers indicate the percent-
ages of splenocytes that fall within the indicated regions. (b) Bar graph showing the per-
centage of clonotypic TCRhiCD4+ splenocytes for mice of the indicated genotypes. Error
bars indicate standard deviation. Differences in cell numbers were considered significant
if t-test P values (two-sample assuming equal variances) were below 0.05. By this crite-
rion, B10.Br TCR mice showed substantially more clonotypic TCRhiCD4+ splenocytes
than all other genotypes, including Aire–/– TCR-insHEL mice. Aire–/– TCR-insHEL mice in
turn showed substantially more clonotypic TCRhi CD4+ splenocytes than B10.Br and
Aire+/+ TCR-insHEL mice. No other significant differences were observed.

staining with a 3A9 TCR clonotype-specific antibody, 1G1231, in the Production of CD4+CD25+ T cells
thymus of TCR-insHEL Aire–/– mice and in a range of littermate con- A longstanding alternative model for explaining tolerance to organ-
trols (Fig. 1). In the thymus of TCR-transgenic animals (lacking specific antigens is the concept of educating self-antigen–specific T
insHEL), T cells bearing the HEL-specific TCR were positively select- cells in the thymus to be exported as regulatory and/or suppressor cells
ed to mature as CD4+CD8– single-positive cells (Fig. 1a). The CD4+ T that prevent autoimmunity in the periphery. CD4+CD25+ cells in the
cells expressed high densities of the HEL-specific TCR on their sur- mouse have been suggested to have this property33–35. Therefore, the
face, and displayed the mature phenotype of down-regulated CD69 production of CD4+CD25+ putative regulatory T cells was also moni-
(activation/developmental marker) and up-regulated receptors (CCR7 tored during thymic development. In TCR transgenic mice without the
or CCR11) for the chemokine CCL19 (ELC)32 (Fig. 1a). Based on these insHEL transgene, only 2% of TCRhiCD4+CD8– T cells were CD25hi
markers, there was no difference in the positive selection of HEL-spe- (Table 1). In TCR-insHEL double-transgenic mice, the proportion of
cific T cells between Aire–/– and the background B10.Br (Aire+/+) TCR CD25+ T cells among TCRhiCD4+CD8– cells increased to 25%. This
transgenic animals (Supplementary Table 1). increase was completely mitigated in Aire–/– TCR-insHEL double-trans-
In the thymus of TCR-insHEL double-transgenic mice that were genic mice, with only 2% of TCRhiCD4+CD8– cells expressing high
either Aire+/+ littermates or B10.BR Aire wild-type animals, few CD4 amounts of CD25 (Table 1). When this subset of T cells was measured
single-positive cells were formed (Fig. 1b and Table 1, see as an absolute number, however, there was no significant difference
Supplementary Table 1 for more details) and of those a minority had between any of the groups. Thus, the increase in percentage of HEL-
matured to express high densities of HEL-specific TCR (Fig. 1b and specific CD4+25+ cells in Aire wild-type animals is secondary to the
Fig. 2), reflecting efficient T cell clonal deletion (negative selection). loss of HEL-specific CD4+25– cells.
Negative selection was also occurring in the immature CD4+8+ double-
positive population, so that the immature population was reduced to Thymus-intrinsic mechanism of Aire-mediated deletion
half that found in TCR animals (Table 1). The small number of CD4+ To determine the cellular requirement for Aire during negative selec-
T cells that mature in the TCR-insHEL thymus carry proportionately tion, chimeric mice were constructed with defective Aire in either the
reduced clonotype TCR and CD3, and ele-
vated CD5, on their cell surface (Fig. 1b).
We have previously shown that these resid- Table 2. The role of Aire in negative selection is intrinsic to the nonhemopoietic
ual cells are anergic based on poor prolifer- compartment
ative response to antigen or TCR Vβ anti- Donor mouse B10.Br TCR+ Aire–/– TCR+
body14. In contrast, in the thymus of Aire–/–
Recipient mouse B10.Br B10.Br HEL+ B10.Br B10.Br HEL+
TCR-insHEL mice, negative selection was
negligible in both the double-positive and Single-positive cells (CD4+CD8–) (%) 27 ± 2 7.8 ± 2 31 ± 8 10 ± 2
immature single-positive subsets. Mature CD4+CD8–1G12hi cells (%) 25 ± 4 4±3 29 ± 8 7±3
CD4+ T cells with high densities of HEL- CD4+CD8–1G12hi CD69– cells (%) 14 ± 2 1.2 ± 0.9 16 ± 5 2.4 ± 0.9
specific TCR, low CD69 and high expres-
sion of CCL19 receptor were produced in Donor mouse B10.Br TCR+
almost normal numbers, with normal cell- Recipient mouse B10.Br B10.Br HEL+ Aire–/– Aire–/– HEL+
surface expression of CD3 and CD5 (Fig.
1d, Fig. 2 and Table 1). This failure of Single-positive cells (CD4+CD8–) (%) 24 ± 2 9.6 ± 2 14 ± 4 13 ± 5
thymic negative selection resulted in CD4+CD8–1G12hi cells (%) 18.7 ± 2 5.1 ± 2 8.6 ± 5 8.7 ± 3
increased export of high-avidity, islet-HEL CD4+CD8–1G12hiCD69– cells (%) 5.9 ± 1 0.8 ± 0.4 2.4 ± 1 1.7 ± 0.6
specific T cells to the peripheral circulation Mean ± s.d.
(Table 1 and Fig. 3).

352 nature immunology • volume 4 no 4 • april 2003 • www.nature.com/natureimmunology


hematopoietic or nonhematopoietic compartments. Analysis of thymic
composition after 6 weeks of reconstitution indicated that Aire–/– HEL-
specific T cells were deleted normally in the thymus of Aire+/+ insHEL
mice (Table 2 and Supplementary Table 2). In contrast, Aire+/+ HEL-
specific T cells were not deleted in Aire–/– insHEL mice, demonstrating
that the trait is intrinsic to the nonhematopoietic compartment (Table 2
and Supplementary Table 2). These data correlate well with the
restriction of thymic Aire expression to the nonhematopoietic lin-
© 2003 Nature Publishing Group http://www.nature.com/natureimmunology
eages23,36, and the thymic-intrinsic induction of autoimmunity in
chimeric mice using Aire–/– mice as recipients for Aire+/+ bone marrow26.
Noticeably, in the Aire–/– recipients without the insHEL transgene, pos-
itive selection of HEL-specific T cells seemed to be reduced in Aire–/–
mice at this early time point, a trait which was not observed in
nonchimeric Aire–/– TCR transgenic mice. This may, therefore, repre-
sent a quantitative impediment in normal T cell maturation in Aire–/–
thymus, which is apparent during the wave of reconstitution, but not
once homeostasis is achieved.
Discussion
We have demonstrated here the defect in thymic negative selection of
high-avidity, organ-specific T cells in Aire–/– mice, which is one of the
first instances in which the cellular mechanism for a human autoim-
mune disease has been identified, and is the first evidence directly
connecting human autoimmune disease with defects in thymic dele-
tion. This result was unexpected because there is a body of evidence
supporting the view that tolerance to organ-specific antigens must
depend, for the most part, on peripheral mechanisms acting on T cells
after they have left the thymus, rather than on central tolerance dur-
ing T cell development, due to the low expression of organ-specific
antigens in the thymus37. Moreover, we have previously considered
that the thymic deletion to insHEL might be caused by circulating
HEL, which is cleaved from islet β cells and is present at nanomolar
serum concentrations comparable to those of insulin itself14. However,
this idea is changing due to the unexpected discovery that the thymus
is capable of promiscuous transcription of a wide range of ‘organ-
specific’ genes17. Originally observed in transgenes with organ-spe-
cific promoters (including the rat insulin promoter used here16), arti-
factual promoter ‘leakiness’ was excluded with the observation that
many organ-specific endogenous genes are also expressed in the thy-
mus38. Ectopic expression has been proposed to reflect a specialized
physiological process, with rare PAE cells in the thymic medullary
epithelium expressing a range of peripheral antigens17. Anderson et
al.26 have recently shown that some ectopic mRNA expression is
dependent on Aire, notably a defect in insulin mRNA transcription in
Aire–/– medullary stromal cells. These data, however, do not demon-
strate a biological role for Aire-dependent mRNAs, and the mecha-
nism of any tolerance induced by Aire is not discussed. In the insHEL
transgenic mice studied here, membrane-bound HEL protein levels
were too low to detect on thymic stromal cells in Aire wild-type mice
(D. Gray and A.L., data not shown). Thus, it was not possible to deter-
mine the direct effect of the absence of Aire on HEL expression in the
thymus at the protein level. However, we conclude from the abroga-
tion of HEL-specific negative selection in Aire–/– mice that the rat
insulin promoter drives Aire-dependent ectopic expression of self
antigen in a form capable of efficiently triggering T cell elimination,
despite the undetectable protein expression. Mixed bone-marrow
chimera experiments that dilute the precursor frequencies of TCR
transgenic cells indicate that the efficient deletion observed in TCR-
insHEL thymus holds true at physiological precursor frequencies
(data not shown).
www.nature.com/natureimmunology • april 2003
A RTICLES
The defective thymic deletion of organ-specific CD4+ T cells in
mice with an Aire mutation does not exclude the possibility that other
tolerance mechanisms may also depend on Aire. The answer will only
be found with greater understanding of the mechanism by which Aire-
dependent ectopic expression results in central tolerance. Although the
mechanism is efficient for deleting 3A9 T cells, which recognize a
dominant peptide epitope of HEL, TCR-transgenic T cells specific for
a subdominant epitope of HEL are not deleted in the same insHEL
transgenic line, even though these cells are still capable of responding
to antigen levels expressed in the periphery and islets (S. Hartley, M.
Neighbors and A. O’Garra, personal communication). Thymic dele-
tion of high-affinity T cells recognizing dominant epitopes may be suf-
ficient for tolerance, by reducing the autoreactive precursor frequency
to unresponsive levels. However, several other possibilities must be
considered. For example, the thymic induction of anergic T cells with
lower surface TCR-CD3 expression and elevated CD5, observed in the
thymus of TCR-insHEL Aire wild-type mice, may be of functional
importance. Not only does anergy prevent autoreactive T cells from
responding39, but when precursor frequencies are reduced by thymic
deletion, residual anergic cells may be able to induce tolerance in the
periphery40 via linked suppression (tolerogenic antigen-presenting
cells)41,42, competition for stimulatory cytokines43, or direct T cell–T
cell interactions10. In one TCR transgenic model, thymic expression of
antigen induces the development of anergic autoreactive cells that cor-
respond phenotypically and functionally to the subset of thymus-
dependent CD4+25+ regulatory T cells present in normal lymphocyte
repertoires33–35. However, in the model used here, Aire-dependent
ectopic expression of insHEL did not alter the number of anti-HEL
TCRhiCD4+CD25+ T cells in the thymus, although the efficient deletion
of TCRhiCD25– T cells resulted in large shift in the ratio of autoreac-
tive to CD4+CD25+ regulatory T cells in TCR-insHEL mice that was
abrogated in Aire-deficient mice. Because CD4+CD25+ regulatory T
cells seem to exist in a semi-anergic state44, Aire-dependent deletion
and anergy may simultaneously lower the frequency of the most
strongly organ-reactive clones and enrich the remaining repertoire
with organ-reactive anergic T cells that have a capacity to suppress the
remaining naive T cells.
The failure of thymic deletion provides a clearer understanding for
autoimmune polyendocrinopathy pathology by demonstrating that Aire
is essential for a highly efficient and specialized mechanism of thymic
T cell deletion. This result is an important step for the field because it
connects human immunological tolerance with the large body of work
on thymic negative selection in transgenic mice. Bridging human
autoimmunity and tolerance mechanisms in the mouse also has mech-
anistic implications for more common and genetically complex autoim-
mune diseases. Comparable lesions in thymic deletion may be pro-
duced by genetic variants acting at various steps in the process of
expression, presentation and recognition of organ-specific antigens by
maturing thymocytes. For example, the type 1 diabetes susceptibility
allele of the human insulin gene diminishes expression in the thy-
mus45,46, and the diabetes-prone NOD mouse strain contains undeter-
mined genetic lesions that result in a T cell intrinsic resistance to
thymic deletion27,47. Multiple sclerosis may also result from a defect in
this process, as susceptibility to experimental autoimmune
encephalomyelitis in the SJL/J mouse was found to correlate with the
thymic expression of a splice variant of the myelin-sheath protein, pro-
teolipid protein (PLP), which lacks the encephalitogenic epitope48,49.
Identifying other genes and molecules in the pathway of thymic cen-
soring may therefore illuminate a common causality for a diverse range
of organ-specific autoimmune diseases.
• volume 4 no 4 • nature immunology 353
 A RTICLES
Methods
Mice. The 3A9 TCR transgenic28 and ILK-3 Ins-HEL transgenic14 mice produced in
C57BL/6J mice were backcrossed >7 generations to B10.Br/SgSnJ (JAX). (B6 × 129/Sv)F2
Aire+/– mice23 were backcrossed to TCR-insHEL B10.Br mice for two generations.
Backcross offspring were typed for TCR, insHEL, H-214 and Aire23 by PCR, and H-2k Aire+/–
mice of appropriate TCR-insHEL genotype were intercrossed to produce Aire–/– H-2k TCR-
insHEL mice. Experimental mice were twice confirmed for genotype by PCR. Radiation
chimeras were constructed as described14. Recipients were analyzed ∼6 weeks after recon-
stitution. Animals were cared for and used in accordance with principles outlined by the
Australian National University Animal Experimentation Ethics Committee and the current
© 2003 Nature Publishing Group http://www.nature.com/natureimmunology
Australian Code of Practice for the Care and Use of Animals for Scientific Purposes.
Flow cytometry. We analyzed 6- to 10-week-old mice (or 6 week post-reconstitution
chimeric mice) as described27,47 using the following antibodies: 1G12 anti-clonotype31 (gift
of E. Unanue and D. Peterson, Washington University, St. Louis, MO) culture supernatant
followed by rat anti-mouse IgG1 allo-phycocyanin; anti-CD8α-PerCP; anti-CD4-FITC or
PE, anti-CD3-PE, anti-CD69-PE (all from BD PharMingen, San Jose, CA), anti-CD5-FITC
and anti-CD25-PE (both from Caltag, Burlingame, CA). Staining for CCL19/ELC receptors
was performed as described50 using CCL19-Fc fusion protein and goat polyclonal anti-
human IgG Fcγ-PE (Jackson Immunoresearch, West Grove, PA).
Note: Supplementary information is available on the Nature Immunology website.
Acknowledgments
We thank D. Gray and R. Boyd for thymic HEL measurement; J. Cyster, E. Unanue and D.
Peterson for staining reagents; the staff of Medical Genome Centre for curating the
mouse colony; and A. Murtagh, S. Ewing and S.Ward for genotyping.This work was sup-
ported by a grant from the Juvenile Diabetes Research Foundation and NHMRC.
Competing interests statement
The authors declare that they have no competing financial interests.
Received 9 December 2002; accepted 11 February 2003.
1. Ohashi, P.S. et al. Ablation of “tolerance” and induction of diabetes by virus infection in viral antigen
transgenic mice. Cell 65, 305–317 (1991).
2. Katz, J.D.,Wang, B., Haskins, K., Benoist, C. & Mathis, D. Following a diabetogenic T cell from genesis
through pathogenesis. Cell 74, 1089–1100 (1993).
3. Goverman, J.Tolerance and autoimmunity in TCR transgenic mice specific for myelin basic protein.
Immunol. Rev. 169, 147–159 (1999).
4. Scott, B. et al. A role for non-MHC genetic polymorphism in susceptibility to spontaneous autoim-
munity. Immunity 1, 73–83 (1994).
5. Lafaille, J.J., Nagashima, K., Katsuki, M. & Tonegawa, S. High incidence of spontaneous autoimmune
encephalomyelitis in immunodeficient anti-myelin basic protein T cell receptor transgenic mice. Cell
78, 399–408 (1994).
6. Waldner, H.,Whitters, M.J., Sobel, R.A., Collins, M. & Kuchroo,V.K. Fulminant spontaneous autoimmu-
nity of the central nervous system in mice transgenic for the myelin proteolipid protein-specific T
cell receptor. Proc. Natl. Acad. Sci. USA 97, 3412–3417 (2000).
7. Hernandez, J., Aung, S., Redmond,W.L. & Sherman, L.A. Phenotypic and functional analysis of CD8(+)
T cells undergoing peripheral deletion in response to cross-presentation of self-antigen. J. Exp. Med.
194, 707–717 (2001).
8. Lambolez, F., Jooss, K.,Vasseur, F. & Sarukhan, A.Tolerance induction to self antigens by peripheral
dendritic cells. Eur. J. Immunol. 32, 2588–2597 (2002).
9. Greenwald, R.J., Boussiotis,V.A., Lorsbach, R.B., Abbas, A.K. & Sharpe, A.H. CTLA-4 regulates induc-
tion of anergy in vivo. Immunity 14, 145–155 (2001).
10. Shevach, E. CD4+CD25+ suppressor T cells: more questions than answers. Nat. Rev. Immunol. 2,
389–400 (2002).
11. Le Douarin, N. et al. Evidence for a thymus-dependent form of tolerance that is not based on elimi-
nation or anergy of reactive T cells. Immunol. Rev. 149, 35–53 (1996).
12. Hammerling, G.J. et al. Non-deletional mechanisms of peripheral and central tolerance: studies with
transgenic mice with tissue-specific expression of a foreign MHC class I antigen. Immunol. Rev. 122,
47–67 (1991).
13. Hoffmann, M.W., Heath,W.R., Ruschmeyer, D. & Miller, J.F. Deletion of high-avidity T cells by thymic
epithelium. Proc. Natl. Acad. Sci. USA 92, 9851–9855 (1995).
14. Akkaraju, S. et al. A range of CD4 T cell tolerance: partial inactivation to organ-specific antigen
allows nondestructive thyroiditis or insulitis. Immunity 7, 255–271 (1997).
15. Klein, L., Klein,T., Ruther, U. & Kyewski, B. CD4 T cell tolerance to human C-reactive protein, an
inducible serum protein, is mediated by medullary thymic epithelium. J. Exp. Med. 188, 5–16 (1998).
354 nature immunology • volume 4 no 4 • april 2003
16. Jolicoeur, C., Hanahan, D. & Smith, K.M.T-cell tolerance toward a transgenic β-cell antigen and tran-
scription of endogenous pancreatic genes in thymus. Proc. Natl. Acad. Sci. USA 91, 6707–6711 (1994).
17. Kyewski, B., Derbinski, J., Gotter, J. & Klein, L. Promiscuous gene expression and central T-cell toler-
ance: more than meets the eye. Trends Immunol. 23, 364–371 (2002).
18. Betterle, C., Greggio, N.A. & Volpato, M. Clinical review 93: autoimmune polyglandular syndrome
type 1. J. Clin. Endocrinol. Metab. 83, 1049–1055 (1998).
19. Nagamine, K. et al. Positional cloning of the APECED gene. Nat. Genet. 17, 393–398 (1997).
20. The Finnish-German APECED Consortium. An autoimmune disease, APECED, caused by mutations
in a novel gene featuring two PHD-type zinc-finger domains. Nat. Genet. 17, 399–403 (1997).
21. Pitkanen, J. et al.The autoimmune regulator protein has transcriptional transactivating properties and
interacts with the common coactivator CREB-binding protein. J. Biol. Chem. 275, 16802–16809
(2000).
22. Bjorses, P. et al. Mutations in the AIRE gene: effects on subcellular location and transactivation func-
tion of the autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy protein. Am. J. Hum.
Genet. 66, 378–392 (2000).
23. Ramsey, C. et al. Aire deficient mice develop multiple features of APECED phenotype and show
altered immune response. Hum. Mol. Genet. 11, 397–409 (2002).
24. Heino, M. et al. Autoimmune regulator is expressed in the cells regulating immune tolerance in thy-
mus medulla. Biochem. Biophys. Res. Commun. 257, 821–825 (1999).
25. Bjorses, P. et al. Localization of the APECED protein in distinct nuclear structures. Hum. Mol. Genet.
8, 259–266 (1999).
26. Anderson, M.S. et al. Projection of an immunological self-shadow within the thymus by the Aire pro-
tein. Science 298, 1395–1401 (2002).
27. Lesage, S. et al. Failure to censor forbidden clone of CD4 T cells in autoimmune diabetes. J. Exp. Med.
196, 1175–1188 (2002).
28. Ho,W.Y., Cooke, M.P., Goodnow, C.C. & Davis, M.M. Resting and anergic B cells are defective in
CD28-dependent costimulation of naive CD4+ T cells. J. Exp. Med. 179, 1539–1549 (1994).
29. Adelstein, S. et al. Induction of self-tolerance in T cells but not B cells of transgenic mice expressing
little self antigen. Science 251, 1223–1225 (1991).
30. Akkaraju, S., Canaan, K. & Goodnow, C.C. Self-reactive B cells are not eliminated or inactivated by
autoantigen expressed on thyroid epithelial cells. J. Exp. Med. 186, 2005–2012 (1997).
31. Van Parijs, L., Peterson, D.A. & Abbas, A.K.The Fas/Fas ligand pathway and Bcl-2 regulate T cell
responses to model self and foreign antigens. Immunity 8, 265–274 (1998).
32. Ueno,T. et al. Role for CCR7 ligands in the emigration of newly generated T lymphocytes from the
neonatal thymus. Immunity 16, 205–218 (2002).
33. Jordan, M.S. et al.Thymic selection of CD4+CD25+ regulatory T cells induced by an agonist self-pep-
tide. Nat. Immunol. 2, 301–306 (2001).
34. Mason, D. Some quantitative aspects of T-cell repertoire selection: the requirement for regulatory T
cells. Immunol. Rev. 182, 80–88 (2001).
35. Sakaguchi, S. Regulatory T cells: key controllers of immunologic self-tolerance. Cell 101, 455–458
(2000).
36. Heino, M. et al. RNA and protein expression of the murine autoimmune regulator gene (Aire) in
normal, RelB-deficient and in NOD mouse. Eur. J. Immunol. 30, 1884–1893 (2000).
37. Stockinger, B.T lymphocyte tolerance: from thymic deletion to peripheral control mechanisms. Adv.
Immunol. 71, 229–265 (1999).
38. Hanahan, D. Peripheral-antigen-expressing cells in thymic medulla: factors in self-tolerance and
autoimmunity. Curr. Opin. Immunol. 10, 656–662 (1998).
39. Schwartz, R.H.T cell anergy. Annu. Rev. Immunol. 21, 305–334 (2003).
40. Jooss, K., Gjata, B., Danos, O., von Boehmer, H. & Sarukhan, A. Regulatory function of in vivo aner-
gized CD4(+) T cells. Proc. Natl. Acad. Sci. USA 98, 8738–8743 (2001).
41. Vendetti, S. et al. Anergic T cells inhibit the antigen-presenting function of dendritic cells. J. Immunol.
165, 1175–1181 (2000).
42. Taams, L.S. et al. Anergic T cells actively suppress T cell responses via the antigen-presenting cell. Eur.
J. Immunol. 28, 2902–2912 (1998).
43. Lombardi, G., Sidhu, S., Batchelor, R. & Lechler, R. Anergic T cells as suppressor cells in vitro. Science
264, 1587–1589 (1994).
44. Sakaguchi, S. et al. Immunologic tolerance maintained by CD25+ CD4+ regulatory T cells: their com-
mon role in controlling autoimmunity, tumor immunity, and transplantation tolerance. Immunol. Rev.
182, 18–32 (2001).
45. Vafiadis, P. et al. Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2
locus. Nat. Genet. 15, 289–292 (1997).
46. Pugliese, A. et al.The insulin gene is transcribed in the human thymus and transcription levels corre-
lated with allelic variation at the INS VNTR-IDDM2 susceptibility locus for type 1 diabetes. Nat.
Genet. 15, 293–297 (1997).
47. Kishimoto, H. & Sprent, J. A defect in central tolerance in NOD mice. Nat. Immunol. 2, 1025–1031
(2001).
48. Klein, L., Klugmann, M., Nave, K.A.,Tuohy,V.K. & Kyewski, B. Shaping of the autoreactive T-cell reper-
toire by a splice variant of self protein expressed in thymic epithelial cells. Nat. Med. 6, 56–61
(2000).
49. Anderson, A.C. et al. High frequency of autoreactive myelin proteolipid protein-specific T cells in the
periphery of naive mice: mechanisms of selection of the self-reactive repertoire. J. Exp. Med. 191,
761–770 (2000).
50. Reif, K. et al. Balanced responsiveness to chemoattractants from adjacent zones determines B-cell
position. Nature 416, 94–99 (2002).
• www.nature.com/natureimmunology