1Department of Pathology, Harvard Medical School, Boston, Massachusetts,
USA. 2These authors contributed equally to this work. Correspondence should
be addressed to D.M. (email@example.com) or C.B. (firstname.lastname@example.org.
Published online 18 June 2009; doi:10.1038/ni.1760/
Foxp3+ regulatory T cells: differentiation,
Markus Feuerer1,2, Jonathan A Hill1,2, Diane Mathis1,2 & Christophe Benoist1,2
Regulatory T cells (Treg cells) characterized by expression of the transcription factor Foxp3 play a key role in immune homeostasis.
Rather than a monomorphic population strictly determined by Foxp3 as a ‘master regulator’, the emerging view is one of Treg cells
as a population with many levels of complexity. Several regulatory factors partake in the control of their transcriptional ‘signature’,
with Foxp3 being a key regulator but insufficient and unnecessary to specify all aspects of the lineage. Distinct subphenotypes of
Foxp3+ Treg cells are found in different anatomical locations. Some subphenotypes specifically control different facets of effector
T cell function and, perhaps surprisingly, share transcriptional control elements with the very cells they regulate. This review will
focus on these novel aspects of Treg cell diversity.
Any biological system involves negative feedback, and it is now recog-
nized that regulatory T cells (Treg cells) play key roles in the maintenance
of lymphoid homeostasis in a number of immune circumstances. These
cells maintain tolerance to self and control autoimmune deviation1,2,
prevent runaway responses to pathogens or allergens, help maintain
a balance with obligate microbial flora3, and facilitate tumors’ escape
from immune monitoring4. Although several distinct lineages may
participate in these functions, an important population was initially
identified in the mouse as CD4+CD25+ or CD4+CD45RB– and was
able to control autoimmunity elicited by thymectomy or lymphopenic
complementation5,6. A firm molecular definition for these cells came
about with the discovery that they express the forkhead–winged helix
transcription factor Foxp3 (refs. 7–10) and that deficiencies in Foxp3
underlie the lymphoproliferation and multiorgan autoimmunity of
scurfy mutant mice and human patients with immunodysregulation
polyendocrinopathy and enteropathy, X-linked (IPEX) syndrome11.
Foxp3+ Treg cells use the αβ T cell antigen receptor (TCR) for anti-
gen recognition and have a broad TCR repertoire similar in size but
largely distinct in composition relative to that of CD4+ conventional
T cells (Tconv cells)12–14. The mechanisms of action of Treg cells are
clearly pleiomorphic, and several modes and mediators of their activ-
ity that are not mutually exclusive have been described, whose relative
importance has yet to be sorted out15. Because of their fundamental
importance for immune function and because of their great potential
for therapeutic modulation, Foxp3+ Treg cells have attracted extraor-
A wide array of mice with conditional knockout of genes and mice
expressing transgenes that report Treg cell existence or function have
been constructed, and these have been the subject of intense genomic,
genetic and epigenetic investigation. More genome-wide transcriptional
profiles have been generated on Treg cells than on any other immune cell
type, which has resulted in the definition of a canonical ‘Treg signature’
that distinguishes Treg cells from Tconv cells, at least in their resting states
in lymphoid organs16–21 (Fig. 1). The Treg signature includes genes over-
expressed or repressed in Treg cells (in a proportion of 2 to 1), genes that
encode proteins with a wide range of cellular locations and functions:
cell surface receptors, signaling kinases and transcription factors. With
bioinformatic treatments that can detect fine variations, up to ~1,500
genes are found to be differentially expressed in Treg cells21, but none
or very few of these differences are absolute; instead, these variations
correspond to quantitative differences between Treg cells and Tconv cells.
Nor are they specific, as almost all transcripts overexpressed in Treg
cells can also be found in non–T cell lineages. A fraction of these Treg
signature genes have also been identified in chromatin immunoprecipi-
tation experiments with antibody to Foxp3 (refs. 22,23). However, the
overlap between differentially expressed genes and Foxp3-bound genes
is not absolute, in part because of the technical limitations of chromatin
immunoprecipitation and in part because Foxp3 does not control all
aspects of the Treg signature (discussed below).
Differentiation and specification in the thymus
Two origins have been described for Foxp3+ cells, whose numeric and
functional importance remain a matter of debate. The first is the thy-
mus, where Foxp3+ cells are generated roughly in sync with positive
selection of conventional CD4+ T cells. The second is the periphery,
where a number of triggers induce the expression of Foxp3 in Tconv
cells; we will refer to this event as ‘conversion’, avoiding the ‘natural
versus adaptive’ terminology, which could lead to the mistaken belief
that some Treg cells would be unnatural or innate (which is untrue, as
all Treg cells express rearranged antigen receptors that define adaptive
lymphocytes). We will first deal with the establishment and transcrip-
tional control of the thymus-derived population before considering the
generation and function of converted Treg cells.
nature immunology volume 10 number 7 july 2009
© 2009 Nature America, Inc. All rights reserved.
How immature thymocytes are selected into the Treg cell alternative
lineage remains an unresolved question. There is strong genetic vari-
ability in their selection and homeostasis, which is perhaps surpris-
ing for a population of such importance in immunoregulation24–26;
Foxp3+ cells are first detected among immature CD4+CD8+ double-
positive cells, but the majority are probably generated from cells that
already underwent positive selection16,27, mainly along the CD4+
single-positive lineage. CD8+ Foxp3+ cells are normally very rare but
can be observed in experimental conditions of thwarted selection of
the CD4+ lineage16,28,29 and perhaps in human patients treated with
antibody to CD3 (ref. 30). Cells entering the Treg lineage can thus be
thought of as cells that have already committed to maturation and
differentiation along the main CD4+ or CD8+ lineages. Although the
rare Foxp3+ CD4+CD8+ double-positive cells have yet to be profiled, it
is clear that the transcriptional Treg signature is established very early
on, with its main characteristics being already present in CD4+ single-
positive cells21. Positive selection of Treg cells requires TCR—major
histocompatibility complex (MHC) molecular interaction, as for Tconv
cells16 but with a stronger dependence on costimulatory signals through
CD28 (refs. 31,32). The different TCR repertoires of Treg cells and Tconv
cells indicate that commitment to the Treg lineage must directly or indi-
rectly reflect differential signals received from the TCR. Engagement
by agonist ligands favors the selection of Treg cells either by inducing
differentiation along the lineage, as observed in transgenic systems33–36,
or because Foxp3+ cells are inherently more resistant to clonal dele-
tion37–39. It would be an oversimplification, however, to extrapolate
that all Treg cells are necessarily self-reactive. Not only are the data on
self-reactivity of Treg cells of normal mice contradictory and not defini-
tive40–42, but also the significant proportion of TCR sequences that are
shared by thymic Treg cell and Tconv cell repertoires12–14 indicates that
many Treg cells are no more self-reactive than are Tconv cells. Rather
than a sharp self-reactive versus non–self-reactive dichotomy distin-
guishing Treg from Tconv cells, it is probably more useful to consider a
probabilistic determinism in which each TCR
has, in a given MHC environment, a distinct
probability to promote commitment along the
Treg lineage, with self-reactivity being one but
not the only determinant. Further along this
line, it has been shown that the proportion of
cells that mature into the Treg lineage is strik-
ingly dependent on the precursor frequency
of a given clone (with very little selection
occurring above a frequency of 1%)43. This
phenomenon could not be accounted for by a
helper effect of additional polyclonal cells but
most likely cannot be accounted for by intra-
clonal competition for MHC-peptide com-
plexes, much as T cells of the same specificity
compete during antigen- or lymphopenia-
driven population expansion. These obser-
vations explained the puzzling mystery that
all MHC class II–restricted TCR–transgenic
mice made monoclonal by crossing onto
the recombination-activating gene–deficient
background have essentially no Foxp3+ cells
and will mandate a reexamination of past data.
Limiting niches have been reported for posi-
tive selection of conventional repertoires44,45,
but the niche size for a given Treg TCR speci-
ficity seems one or two orders of magnitude
The Hsieh and Farrar groups have described a two-step Treg cell
differentiation process in which a Foxp3– CD25hi population already
enriched in TCR sequences ‘preferentially’ found in mature Treg cells is
the first intermediate. Exposure to interleukin 2 (IL-2) can then convert
these intermediates into fully differentiated CD25+Foxp3+ cells46,47.
This importance of IL-2 in eliciting Foxp3 expression is consistent with
the profound Treg cell defects in mice lacking the IL-2 receptor or the
IL-2 signal transducers Jak3 or STAT5 (ref. 2). However, most evidence
indicates that transforming growth factor-β (TGF-β) is not required
for thymic selection of Treg cells as it is for their later homeostasis in
the periphery48–50, except perhaps in the neonatal period, as T cells
devoid of TGF-β receptor I show a slight delay in the appearance of
thymic Treg cells51.
How these differential TCR signals are translated into ‘preferential’
Treg cell commitment is beginning to be better understood. Several
milestones have been put down that etch a putative map of how the
differential TCR signals are channeled through signaling pathways to
induce Treg cell differentiation (Fig. 2). Activation of the transcription
factor NF-κB pathway seems particularly important for Treg cell differ-
entiation, more so than for normal T cells, as deficits in several elements
that link the TCR to NF-κB have been proven highly deleterious for Treg
cell development. Mice with conditional knockout of PKC-θ, Bcl-10,
CARMA1 or IKK2 have defective Treg cell selection52–57. These four
molecules draw a fairly clear path from the TCR to NF-κB activation.
The MAPK kinase kinase TAK1 (also called Map3k7) is also essential for
Treg cell selection58,59, although this observation is harder to pinpoint
on signaling maps, as TAK1 is involved in cytokine signaling (TGF-β,
IL-2, IL-15) as well as TCR signaling. However, signaling through the
kinase Akt pathway has a negative effect on Treg cell thymic selection,
as constitutively active Akt impairs the thymic differentiation of Treg
cells, as well as their conversion by TGF-β60,61, consistent with a positive
effect of the kinase mTOR inhibitor rapamycin on Treg cell selection
and population expansion60–65. These effects are probably related to
volume 10 number 7 july 2009 nature immunology
KInase or phosphatase
DNA binding or
GTP or ATP binding
Figure 1 Treg cell signature genes and their cellular localization. The most differentially expressed genes
from a consensus Treg signature21, either overexpressed (red) or underexpressed (blue) in resting Treg
cells of spleen or lymph node relative to Tconv cells. Gene products are grouped according to cellular
localization, and their putative functions are identified by symbols (key).
© 2009 Nature America, Inc. All rights reserved.
the enhanced induction of Foxp3 and corresponding dearth of effector
cytokines that occur after TCR stimulation of mature T cells lacking
mTOR. This activity seems attributable to the TORC2 complex66. Thus,
it is possible that the TCR signals that promote selection into the Treg
cell lineage are those that elicit a particular balance of transduction
along the NF-κB and Akt-mTOR pathways (Fig. 2).
How these signals are then translated into specific activation of Foxp3
and other controllers of the Treg signature remains mysterious. A num-
ber of regulatory factors and pathways affect the activity of the Foxp3
locus67, but most of them are ubiquitous effectors of cellular activation,
alone or in combination (such as AP1, NFAT and CREB).
Foxp3 as a lineage-specification factor?
From the differential inputs described above, how is the Treg cell sig-
nature established and which transcriptional regulators forge it? After
the discovery of mutations in the gene encoding the transcription
regulator Foxp3 as the root of lymphoproliferative and autoimmune
disease in scurfy mice and patients with IPEX and the results of early
transduction experiments8–10, Foxp3 was seen as the ‘master regulator’
of the Treg cell lineage, with its presence being necessary and sufficient
to specify their phenotype and function. This dogma is still represented
in many reviews and in Introductions to primary articles. Yet several
arguments have progressively accumulated to erode this view of Foxp3
as the unique specification factor of the lineage21,68.
First was the description of ‘wannabe’ Treg cells in the thymus by
the Rudensky and Chatila laboratories20,69: in Foxp3–green fluorescent
protein reporter mice in which the green fluorescent protein insert
destroys the encoded Foxp3 protein, a substantial number of cells have
several of the characteristics of Treg cells, including transcriptional
activity at the Foxp3 locus, high expression of the majority of Treg cell
signature genes (including canonical genes such as Il2ra, Nrp1, Ctla4
and Icos) and low expression of Il2 (the last being of particular inter-
est because suppression of IL-2 had been thought to be a direct and
unique result of Foxp3 action). However, these cells were unstable and
exerted no suppressive activity (an interpretation confounded by the
fact these ‘wannabe’ cells themselves adopted effector characteristics).
This observation was also consistent with the existence of Treg-like cells
in some patients with IPEX70. Similarly, it proved possible to elicit a
substantial portion of the Treg cell signature in cells devoid of Foxp3
(for example, in TGF-β-treated scurfy mutants or by homeostatic con-
version in vivo of Foxp3-null cells; refs. 21,71; Fig. 3). Second, careful
analysis of cells in which Foxp3 was expressed by direct transduction,
or by induced conversion (for example, in vitro culture with TGF-β
with or without retinoic acid or in vivo exposure to agonist or in vivo
homeostatic expansion) showed that Foxp3 could restore at most about
one third of the Treg cell signature transcripts19,21,71. In addition, the
functional efficacy of Foxp3-transduced or TGF-β-converted cells has
variably ranged from highly efficacious to largely inactive, for reasons
that remain puzzling although probably related to the stability of Foxp3
expression in TGF-β-converted Treg cells67 (discussed below). These
results indicate that expression of Foxp3 alone does not always suf-
fice for a suppressor phenotype9,10,21,72–74. Thus, Foxp3 seems neither
absolutely necessary nor uniformly sufficient to specify many aspects
of the Treg cell phenotype.
What factors in addition to Foxp3 control the Treg cell signature? A
sizeable fraction probably originates from IL-2 through STAT5 (refs.
21,75), consistent with the two-step model for Treg cell selection, in
which IL-2 plays a central role. Bioinformatic meta-analyses of Treg cell
datasets demonstrating the existence of a group of genes coregulated
with Foxp3 but not induced directly by it suggested the presence of a
higher-order regulatory network21. In this alternative hypothesis, Foxp3
would serve as an important activator or suppressor of a set of genes
(some of which are essential for suppressor function) but would be
complemented by other transcriptional regulators that control their
own set of transcripts in the Treg cell signature. These controls can be
complementary and synergistic, as a given Treg cell signature gene can be
activated by several pathways (for example, CD103 responds to Foxp3
as well as the combination of IL-2–STAT5 and TGF-β independently
of Foxp3). To use a political metaphor, Foxp3 is a primus inter pares (a
member of an oligarchy) rather than a dictatorial master regulator.
Converted Treg cells
As mentioned above, naive Tconv cells can be induced to express Foxp3
by a variety of means (for example, within 2–4 days of activation in
the presence of IL-2 and TGF-β in vitro76,77; within 8–14 days of expo-
sure in vivo to subimmunogenic agonist peptide delivered by osmotic
minipumps or peptide coupled to antibody to DEC205 in transgenic
systems78,79 or polyclonal systems80; after exposure to antigen delivered
through mucosal surfaces81–83; or as a result of lymphopenia-driven
homeostatic proliferation71,84,85). In theory, conversion is an attrac-
tive mechanism, as it allows lymphocyte pools to adapt to immuno-
genic conditions, to dampen an overactive acute inflammation or to
curtail the response to a chronic unresolved challenge. Of note, this
nature immunology volume 10 number 7 july 2009
α α αα
mTORC1mTORC1 mTORC2 mTORC2
IκB IκB IκB
Figure 2 Differential signaling induces or inhibits Treg cell differentiation.
engagement of the NF-κB pathway ‘downstream’ of the TCr and of iL-2–
STAT5 promotes Treg cell differentiation, whereas activation of the Akt-mTOr
arm inhibits this, as suggested by the gene knockouts that diminish (red) or
increase (blue) commitment to the Treg lineage.
© 2009 Nature America, Inc. All rights reserved.
concept represents a departure from the paradigm of clonal selection
that has served immunology well for several decades; this departure
is not truly necessary, as the breadth of the Treg cell TCR repertoire
as it emerges from the thymus can certainly provide Treg cells reac-
tive against any given antigen-MHC complex, these antigen-specific
Treg cells being amplified in situ just as Tconv cells (Treg cells actually
divide more in vivo than Tconv cells, contrary to their anergic activity
in vitro36,86,87). Another view, consistent with the instability of Foxp3
expression observed in some of these conversion settings (discussed
below) is that transiently eliciting inhibitory functions may be a way of
quickly ‘fine-tuning’ the first steps of a local immune response.
The observation that conversion can occur during experimental
manipulation leaves open the question of the true contribution of
peripheral conversion to the overall Treg cell pool and whether this is a
focused response occurring at specific inflammatory locations. There
has been a tendency in the literature to interpret observations of local
Treg cell accumulation as reflecting conversion from Tconv cells, rather
than simple migration, retention and proliferation of antigen-specific
Treg cells, but actual evidence for either is often missing. The question is
of heuristic importance (are Treg cells a distinct lineage or one of several
states into which naive CD4+ T cells can differentiate?) and practical
importance (can conversion be a therapeutic target?). In nonimmunized
and nonlymphopenic mice, the CDR3 sequences of TCRs expressed by
Treg cells isolated from peripheral lymphoid organs largely resemble
those of thymic Treg cells, with no peripheral accentuation of the over-
lap between repertoires that would result from conversion12–14,88; this
finding suggests that the global contribution of converted Treg cells in
lymph nodes and spleen is limited. A more focused analysis using TCR
sequences as ‘barcodes’ to look for evidence of conversion in a setting
of autoantigen recognition, a priori more favorable to detect conver-
sion events, also failed to bring evidence for any substantial numeric
contribution89. Similarly, Treg cells are found in brain inflammatory
lesions in mice with experimental autoimmune encephalomyelitis as a
result of the migration of thymus-derived Treg cells rather than conver-
sion90. In infectious settings, the available evidence often points to the
recruitment and expansion of antigen-specific Treg cell populations,
rather than conversion91,92. An interesting handle on the question may
be provided by mice lacking CARMA1 (ref. 54). As discussed above,
thymic selection of the Foxp3+ lineage is profoundly deficient in these
mice, but Foxp3 can be very effectively induced in their mature Tconv
cells by exposure to IL-2 and TGF-β in vitro. Interestingly, the lamina
propria of CARMA-1 deficient
mice shows a substantial con-
tingent of Treg cells (~37% of
a normal pool), far more than
in mesenteric or other lymph
nodes (8–3%). Although the
possibility of localized expan-
sion of rare thymic precursors
cannot be ruled out, this dis-
tribution could be interpreted
to reflect peripheral conver-
sion induced by TGF-β more
uniquely in the gut-associated
tissue than in spleen or other
From a functional stand-
point and with the exception
of the variable results obtained
with the Treg cells induced by
TGF-β discussed above, cells
converted in vivo can be functionally quite effective71,78,83,84. In the
DO11.10 system, Treg cells elicited by exposure to antigen through the
gut can protect from airway inflammation83. Foxp3+ cells converted
from Foxp3– precursors in conditions of homeostatic expansion are as
effective as resting Treg cells in protection against colitis elicited by trans-
fer of naive Tconv cells into a lymphopenic host; indeed, these converted
Treg cells function even more effectively when combined with resting
lymph node Treg cells, which suggests that they brought a complemen-
tary phenotype or an enriched antigenic specificity71. This observation
is compatible with the notion that these ‘neo–Treg cells’, generated in
lymphopenic conditions, are particularly adept at regulating immu-
nopathology occurring in precisely the same triggering conditions of
the lymphopenic host.
From a genomic standpoint, converted Treg cells are clearly differ-
ent from thymus-derived Treg cells, as demonstrated for TGF-β-Treg
cells and for Treg cells induced in lymphopenic conditions21,71. In both
instances, only a fraction of the Treg cell signature was elicited (~35%);
although canonical transcripts (Foxp3 and Itgae (encoding CD103))
were expressed, others were not differentially expressed (Il2ra and
Ctla4 for TGF-β-Treg cells, Ikzf4 (Helios) and Itgb8 for homeostati-
cally induced Treg cells). Nor are these fractions similar, and Treg cells
converted in different scenarios each have a subtly different subset of
the entire signature (M. Feuerer et al., unpublished data).
It is unclear what relationship exists between the signaling pathways
that promote the selection of the Treg cell lineage in the thymus and
those that elicit conversion in the periphery. Some appear shared (the
importance of TCR engagement, of IL-2 and of Akt). Some appear
distinct, in particular the effect of TGF-β (apparently dispensable in
the thymus; clearly involved in some but not all instances of peripheral
conversion) or of IL-6 and the transcription factors that control the
differentiation of IL-17-producing cells. The latter point is of particu-
lar interest given observations of shared requirements in conditions
that elicit the differentiation of IL-17-producing T helper cells (TH-17
cells) or conversion to a Foxp3+ phenotype in vitro93,94. Both processes
require TGF-β, but IL-6, by inducing expression of the transcription
factor RORγt, effectively shuts down Foxp3 induction95,96. Interestingly,
RORγt and Foxp3 are both induced during the early phase of a TGF-β-
induced response and physically interact, but Foxp3 wins out by shut-
ting down RORγt97. There is no evidence that a similar interaction
occurs between Foxp3 and RORγt in the thymus (for example, RORγt-
deficient mice have normal thymic Treg cells), but might other members
volume 10 number 7 july 2009 nature immunology
Figure 3 Different segments of the Treg signature appear in different contexts. Heat map of the transcripts of a
consensus Treg signature, normalized to the expression in splenic Tconv cells and Treg cells (0 and 1, respectively,
presented as hierarchical clustering). Lymph node (LN) Treg cells have essentially identical profiles, but only a
fraction of the signature is present in Treg cells from adipose tissue or after in vitro conversion with TGF−β. Most of
the signature transcripts acquired after TGF-β-induced conversion are Foxp3 independent, as they are also present in
cultures of Foxp3-deficient scurfy mutant cells (top).
© 2009 Nature America, Inc. All rights reserved.
of the large family of nuclear receptors play a corresponding role?
Conversely, there is also emerging evidence suggesting that the Treg
cell transcriptional programs are not necessarily permanently etched,
but that the Foxp3 transcriptional cassette may be expressed in a revers-
ible manner, transiently or for more extended periods of time (refs.
67,98,99 and J. Bluestone, personal communication). Instability was
linked to the differential stability of Foxp3 expression as a function of
epigenetic changes at the Foxp3 locus67, and it may confer additional
flexibility to the application of regulatory functions. In addition, in the
realm of the Treg cell versus TH-17 cell relationships mentioned above, it
is interesting to note that strong IL-17-inducing conditions can elicit a
shutdown of Foxp3 and induction of IL-17 production from a fraction
of outwardly established Foxp3+ cells96,100.
Functional subphenotypes of Foxp3+ T cells
Soon after the original description of Foxp3 in CD25+CD4+ T cells,
subsets of this population were identified by differential expression
of cell surface markers. Several of these subsets probably correspond
to markers of activation or memory that, as for activated Tconv cells,
allow them to home to locations other than the secondary lymphoid
organs. One of the best characterized examples is the integrin αEβ7
(CD103), which binds E-cadherin. It is typically expressed on 20–30%
of Foxp3+ cells in secondary lymphoid organs and on a higher percent-
age of Treg cells in tissues such as the lung, skin and lamina propria of
the gut17,101,102. The functional relevance of CD103 expression in the
Treg cell population is highlighted by the greater potential of Treg cells
to access and be retained in peripheral tissues during infection or acute
inflammatory insults17,91. CD103 is under multifactorial and complex
regulation: it is directly responsive to Foxp3 after retroviral transduc-
tion in vitro but can be induced by TGF-β in a Foxp3-independent
manner (for example, in naive CD4+ T cells from scurfy mice cultured in
vitro with TGF-β21), and it is strongly expressed in thymic derived and
converted Foxp3+ cells after antigen exposure or homeostatic expansion
(refs. 17,29,102,103 and M. Feuerer, J. Hill, D. Mathis and C. Benoist,
unpublished data). The ‘activated-memory’ CD103+ Treg cell subset can
be further subdivided by the expression of markers typical of natural
killer cells, such as KLRG1, CD49b and CD38 (refs. 29,104), whose
functional relevance is unclear today.
Other Treg cell subsets, in contrast, correlate with specific tissue local-
izations. For instance, the chemokine receptor CCR4 is not expressed
on thymic Treg cells but is found on an unusually high percentage of
extralymphoid Foxp3+ cells in the skin102. CCR4+ Treg cells also appear
in skin draining lymph nodes after subcutaneous immunization, and
their functional relevance is highlighted by the inflammatory mani-
festations that develop in mice in which Ccr4 is conditionally deleted
specifically in Treg cells. CCR4-deficient Treg cells function normally
in in vitro inhibition assays, are competitively fit and are able to con-
trol many of the peripheral tissue manifestations of autoimmunity in
scurfy mice but cannot control inflammation in the skin or lungs due
in part to their impaired ability to migrate or be retained in these tis-
sues. Similar results have also been obtained with mice that lack the
skin-homing receptor α-1,3-fucosyltransferase VII (refs. 105,106).
Thus, the ability of Treg cells to protect against autoimmune damage
in a particular organ requires the ability of Treg cells to home to that
organ; the mere expression of Foxp3 coupled with functional efficacy
in particular in vitro or in vivo assays does not necessarily equate to a
bona fide Treg cell. A distinct population of Foxp3+ cells residing in the
adipose tissue has been described, its presence or absence correlating
with pathological manifestations of obesity and insulin resistance (M.
Feuerer et al., unpublished data). Here again, these cells express only a
subset of the Treg signature (Fig. 3) but also express other transcripts
that may account for their particular location and effector function. It
remains to be resolved whether these Treg cell subpopulations are elic-
ited and acquire their particular characteristics after antigen encounter
(for example, perhaps in contact with particular antigen-presenting
cells or adventitious stimuli at the time of TCR triggering and/or con-
version) or whether a diversity of transcriptional programs are prese-
lected in the thymus, in addition to the bedrock program imparted by
Foxp3, with the cells being later selected through differential homing
and antigenic specificity. In this respect, it is interesting to note that Treg
cells in different lymph nodes have quite different TCR repertoires88.
Although some of these subphenotypes correspond to differential
activation or tissue localization of Treg cells, two reports also indi-
cate that transcriptional submodules in Treg cells are needed for the
regulation of different immune functions and that Treg cells do so by
involving transcriptional control elements from the very cells they are
regulating. This has been shown in the context of the T helper type 1
(TH1) transcription factor T-bet (encoded by Tbx21) and the TH2- and
TH-17-related transcription factor IRF4 (encoded by Irf4)107,108. These
regulators of lineage development have been studied for their ability to
influence cytokine production, but they also help coordinate a much
broader transcriptional program in T cells as well as other immune cell
types109,110. The main outcome of deleting Irf4 uniquely in Foxp3+ Treg
cells is strong overexpression of IL-4 and IL-5 (and to a much more
modest extent IL-17) but not of other cytokines in Tconv cells, and a
massive increase in the production of immunoglobulin G1 and immu-
noglobulin E by B cells; these features are not typical of Foxp3-deficient
mice. Interestingly, the transcriptional program of Irf4-deficient Treg
cells show a small number of changes, many of which affect character-
istic TH2 transcripts such as Maf or Ccr8 or other chemokine receptors
such as those encoded by Ccr2 or Ccr6. Coimmunoprecipitation from
primary cell extracts indicates that IRF4 and Foxp3 are physically asso-
ciated in Treg cells, which suggests that these two transcription factors
might act together to control a Treg cell subsignature. Indeed, combined
analysis of the Treg signature and of the IRF4 ‘footprint’ shows that
many genes controlled by IRF4 belong to the Treg signature but that
IRF4 affects only a limited subsegment of the Treg signature.
A role for T-bet in Treg cells has been demonstrated by another route
in studies of the expression of CXCR3 on Treg cells109. As in conven-
tional TH1 cells, Cxcr3 expression in Treg cells was dependent on T-bet
(clearly not a factor expected to mediate immunoregulation by Treg
cells!), which was induced after TCR stimulation in the context of the
activation of dendritic cells by antibody to CD40, classically a TH1-
inducing condition. T-bet-deficient Treg cells survived less well and were
less effective than their wild-type counterparts at controlling type 1
inflammatory responses in vivo.
Thus, both of the reports discussed above suggest that Treg cells use
the same transcriptional regulators as the cells they restrain to generate
adapted ‘subsignatures’ or transcriptional cassettes needed to control
a particular facet of the immune response. A thorough transcriptional
analysis needs to be performed, but the IRF4 ‘footprint’ in Treg cells
might be expected to be a composite of some of the elements it con-
trols in TH2 cells and of other elements that it uniquely activates in
Treg cells (for example, as a result of combinatorial transactivation by
Why the match between regulator and ‘regulatee’? One scenario is
that shared transcriptional factors would allow the expression of shared
surface molecules responsible for homing of the helper T cell and its
specific Foxp3+ regulator to the same anatomical location, with shared
location explaining the apparent specificity of regulation; such a shared
location could be macroscopic (homing to particular tissues such as the
gut) or microscopic (particular subsections of T cell areas in secondary
nature immunology volume 10 number 7 july 2009
© 2009 Nature America, Inc. All rights reserved.
lymphoid organs). The fact that chemokine receptor expression is one
of the main consequences of Irf4 deletion in Treg cells might support this
hypothesis, consistent with the requirement for chemokine receptors
for Treg cell function102,105,111. Alternatively, competition for a specific
but common survival factor might be involved.
There may also be a parallel here to the role of T-bet in T cells and B
cells112. In T cells, T-bet controls interferon-γ and TH1-like responses,
whereas in B cells it facilitates class switching to the immunoglobu-
lin G2a isotype, precisely the isotype whose use is enhanced by TH1
cells. Although this is perhaps a mere coincidence, this independent
instance of a situation where regulator and regulatee cells share the
same specification factor may open the following line to speculation:
there is advantage in having the same transcriptional cassettes expressed
in both sides of a regulator–regulatee cell (or function) pair because it
ensure coevolution of the same partners.
The studies discussed here underscore the complexity of transcrip-
tional and phenotypic regulation in Treg cells, in which multiple factors
control the bedrock signature as well as the different subfunctions and
subphenotypes. The notion of a unimodal program of T cell differ-
entiation may hold little relevance to the complexity that is inherent
to Treg cell populations in vivo. Clearly, the extent of this diversity and
how stable or interrelated these Treg cells subphenotypes may be is not
known. But this complexity will need to be considered when devising
therapeutic strategies based on Treg cells.
We thank J. Powell, R. Flavell, D. Campbell, J. Lafaille, R. Maizels, V. Kuchroo and
A. Rudensky for discussions and communication of unpublished data. Supported
by the National Institute of Allergy and Infectious Diseases of the US National
Institutes of Health (R01-AI051530), the Juvenile Diabetes Research Foundation
(4-2007-1057), the German Research Foundation (Emmy-Noether Fellowship FE
801/1-1 to M.F.), the Charles A. King Fund (M.F.) and the Canadian Institutes of
Health Research (J.H.).
Published online at http://www.nature.com/natureimmunology/
reprints and permissions information is available online at http://npg.nature.com/
1. Sakaguchi, S. et al. Foxp3+ CD25+ CD4+ natural regulatory T cells in dominant self-
tolerance and autoimmune disease. Immunol. Rev. 212, 8–27 (2006).
Sakaguchi, S., Yamaguchi, T., Nomura, T. & Ono, M. regulatory T cells and immune
tolerance. Cell 133, 775–787 (2008).
Belkaid, Y. & Tarbell, K. regulatory T cells in the control of host-microorganism
interactions. Annu. Rev. Immunol. 27, 551–589 (2009).
Yamaguchi, T. & Sakaguchi, S. regulatory T cells in immune surveillance and treat-
ment of cancer. Semin. Cancer Biol. 16, 115–123 (2006).
Sakaguchi, S., Sakaguchi, N., Asano, M., itoh, M. & Toda, M. immunologic self-
tolerance maintained by activated T cells expressing iL-2 receptor α-chains (CD25).
Breakdown of a single mechanism of self-tolerance causes various autoimmune
diseases. J. Immunol. 155, 1151–1164 (1995).
Saoudi, A., Seddon, B., Heath, v., Fowell, D. & Mason, D. The physiological role of
regulatory T cells in the prevention of autoimmunity: the function of the thymus in the
generation of the regulatory T cell subset. Immunol. Rev. 149, 195–216 (1996).
Brunkow, M.e. et al. Disruption of a new forkhead/winged-helix protein, scurfin,
results in the fatal lymphoproliferative disorder of the scurfy mouse. Nat. Genet. 27,
Khattri, r., Cox, T., Yasayko, S.A. & ramsdell, F. An essential role for Scurfin in
CD4+CD25+ T regulatory cells. Nat. Immunol. 4, 337–342 (2003).
Fontenot, J.D., Gavin, M.A. & rudensky, A.Y. Foxp3 programs the development and
function of CD4+CD25+ regulatory T cells. Nat. Immunol. 4, 330–336 (2003).
10. Hori, S., Nomura, T. & Sakaguchi, S. Control of regulatory T cell development by the
transcription factor Foxp3. Science 299, 1057–1061 (2003).
11. Ziegler, S.F. FOXP3: of mice and men. Annu. Rev. Immunol. 24, 209–226 (2006).
12. Hsieh, C.S., Zheng, Y., Liang, Y., Fontenot, J.D. & rudensky, A.Y. An intersection
between the self-reactive regulatory and nonregulatory T cell receptor repertoires.
Nat. Immunol. 7, 401–410 (2006).
13. Pacholczyk, r., ignatowicz, H., Kraj, P. & ignatowicz, L. Origin and T cell receptor
diversity of Foxp3+CD4+CD25+ T cells. Immunity 25, 249–259 (2006).
14. wong, J. et al. Adaptation of TCr repertoires to self-peptides in regulatory and
nonregulatory CD4+ T cells. J. Immunol. 178, 7032–7041 (2007).
15. vignali, D.A., Collison, L.w. & workman, C.J. How regulatory T cells work. Nat. Rev.
Immunol. 8, 523–532 (2008).
16. Fontenot, J.D. et al. regulatory T cell lineage specification by the forkhead transcrip-
tion factor Foxp3. Immunity 22, 329–341 (2005).
17. Huehn, J. et al. Developmental stage, phenotype, and migration distinguish naive-
and effector/memory-like CD4+ regulatory T cells. J. Exp. Med. 199, 303–313
18. Herman, A.e., Freeman, G.J., Mathis, D. & Benoist, C. CD4+CD25+ T regulatory cells
dependent on iCOS promote regulation of effector cells in the prediabetic lesion.
J. Exp. Med. 199, 1479–1489 (2004).
19. Sugimoto, N. et al. Foxp3-dependent and -independent molecules specific for
CD25+CD4+ natural regulatory T cells revealed by DNA microarray analysis. Int.
Immunol. 18, 1197–1209 (2006).
20. Lin, w. et al. regulatory T cell development in the absence of functional Foxp3. Nat.
Immunol. 8, 359–368 (2007).
21. Hill, J. et al. Foxp3-dependent and independent regulation of the Treg transcriptional
signature. Immunity 25, 693–695 (2007).
22. Zheng, Y. et al. Genome-wide analysis of Foxp3 target genes in developing and mature
regulatory T cells. Nature 445, 936–940 (2007).
23. Marson, A. et al. Foxp3 occupancy and regulation of key target genes during T-cell
stimulation. Nature 445, 931–935 (2007).
24. Brusko, T. et al. No alterations in the frequency of FOXP3+ regulatory T-cells in type
1 diabetes. Diabetes 56, 604–612 (2007).
25. romagnoli, P., Tellier, J. & van Meerwijk, J.P. Genetic control of thymic development
of CD4+CD25+FoxP3+ regulatory T lymphocytes. Eur. J. Immunol. 35, 3525–3532
26. Feuerer, M. et al. enhanced thymic selection of FoxP3+ regulatory T cells in the NOD
mouse model of autoimmune diabetes. Proc. Natl. Acad. Sci. USA 104, 18181–
27. Liston, A. et al. Differentiation of regulatory Foxp3+ T cells in the thymic cortex. Proc.
Natl. Acad. Sci. USA 105, 11903–11908 (2008).
28. Bienvenu, B. et al. Peripheral CD8+CD25+ T lymphocytes from MHC class ii-deficient
mice exhibit regulatory activity. J. Immunol. 175, 246–253 (2005).
29. Stephens, G.L., Andersson, J. & Shevach, e.M. Distinct subsets of FoxP3+ regulatory
T cells participate in the control of immune responses. J. Immunol. 178, 6901–6911
30. Bisikirska, B., Colgan, J., Luban, J., Bluestone, J.A. & Herold, K.C. TCr stimula-
tion with modified anti-CD3 mAb expands CD8+ T cell population and induces
CD8+CD25+ Tregs. J. Clin. Invest. 115, 2904–2913 (2005).
31. Salomon, B. et al. B7/CD28 costimulation is essential for the homeostasis of the
CD4+CD25+ immunoregulatory T cells that control autoimmune diabetes. Immunity
12, 431–440 (2000).
32. Tai, X., Cowan, M., Feigenbaum, L. & Singer, A. CD28 costimulation of developing
thymocytes induces Foxp3 expression and regulatory T cell differentiation indepen-
dently of interleukin 2. Nat. Immunol. 6, 152–162 (2005).
33. Jordan, M.S. et al. Thymic selection of CD4+CD25+ regulatory T cells induced by an
agonist self-peptide. Nat. Immunol. 2, 283–284 (2001).
34. Apostolou, i., Sarukhan, A., Klein, L. & von Boehmer, H. Origin of regulatory T cells
with known specificity for antigen. Nat. Immunol. 3, 756–763 (2002).
35. Kawahata, K. et al. Generation of CD4+CD25+ regulatory T cells from autoreactive
T cells simultaneously with their negative selection in the thymus and from nonau-
toreactive T cells by endogenous TCr expression. J. Immunol. 168, 4399–4405
36. walker, L.S., Chodos, A., eggena, M., Dooms, H. & Abbas, A.K. Antigen-dependent
proliferation of CD4+CD25+ regulatory T cells in vivo. J. Exp. Med. 198, 249–258
37. van Santen, H.M., Benoist, C. & Mathis, D. Number of Treg cells that differentiate
does not increase upon encounter of agonist ligand on thymic epithelial cells. J. Exp.
Med. 200, 1221–1230 (2004).
38. Liston, A., Lesage, S., wilson, J., Peltonen, L. & Goodnow, C.C. Aire regu-
lates negative selection of organ-specific T cells. Nat. Immunol. 4, 350–354
39. Bonasio, r. et al. Clonal deletion of thymocytes by circulating dendritic cells homing
to the thymus. Nat. Immunol. 7, 1092–1100 (2006).
40. Hsieh, C.S. et al. recognition of the peripheral self by naturally arising CD25+CD4+
T cell receptors. Immunity 21, 267–277 (2004).
41. Pacholczyk, r. et al. Nonself-antigens are the cognate specificities of Foxp3+ regula-
tory T cells. Immunity 27, 493–504 (2007).
42. Dipaolo, r.J. & Shevach, e.M. CD4+ T-cell development in a mouse expressing a
transgenic TCr derived from a Treg. Eur. J. Immunol. 39, 234–240 (2009).
43. Bautista, J.L. et al. intraclonal competition limits the fate determination of regulatory
T cells in the thymus. Nat. Immunol. 10, 610–617 (2009).
44. Huesmann, M., Scott, B., Kisielow, P. & von Boehmer, H. Kinetics and efficacy of
positive selection in the thymus of normal and T cell receptor transgenic mice. Cell
66, 533–540 (1991).
45. Merkenschlager, M., Benoist, C. & Mathis, D. evidence for a single-niche model of
positive selection. Proc. Natl. Acad. Sci. USA 91, 11694–11698 (1994).
46. Lio, C.w. & Hsieh, C.S. A two-step process for thymic regulatory T cell development.
Immunity 28, 100–111 (2008).
47. Burchill, M.A. et al. Linked T cell receptor and cytokine signaling govern the develop-
ment of the regulatory T cell repertoire. Immunity 28, 112–121 (2008).
48. Li, M.O., Sanjabi, S. & Flavell, r.A. Transforming growth factor-β controls develop-
ment, homeostasis, and tolerance of T cells by regulatory T cell-dependent and
-independent mechanisms. Immunity 25, 455–471 (2006).
volume 10 number 7 july 2009 nature immunology
© 2009 Nature America, Inc. All rights reserved.
49. Marie, J.C., Letterio, J.J., Gavin, M. & rudensky, A.Y. TGF-β1 maintains suppressor
function and Foxp3 expression in CD4+CD25+ regulatory T cells. J. Exp. Med. 201,
50. Pesu, M. et al. T-cell-expressed proprotein convertase furin is essential for mainte-
nance of peripheral immune tolerance. Nature 455, 246–250 (2008).
51. Liu, Y. et al. A critical function for TGF-β signaling in the development of natural
CD4+CD25+Foxp3+ regulatory T cells. Nat. Immunol. 9, 632–640 (2008).
52. Schmidt-Supprian, M. et al. Differential dependence of CD4+CD25+ regulatory and
natural killer-like T cells on signals leading to NF-κB activation. Proc. Natl. Acad.
Sci. USA 101, 4566–4571 (2004).
53. Gupta, S. et al. Differential requirement of PKC-θ in the development and function
of natural regulatory T cells. Mol. Immunol. 46, 213–224 (2008).
54. Barnes, M.J. et al. Commitment to the regulatory T cell lineage requires CArMA1 in
the thymus but not in the periphery. PLoS Biol. 7, e51 (2009).
55. Molinero, L.L. et al. CArMA1 controls an early checkpoint in the thymic development
of FoxP3+ regulatory T cells. J. Immunol. 182, 6736–6743 (2009).
56. Medoff, B.D. et al. Differential requirement for CArMA1 in agonist-selected T-cell
development. Eur. J. Immunol. 39, 78–84 (2009).
57. Schmidt-Supprian, M. et al. Mature T cells depend on signaling through the iKK
complex. Immunity 19, 377–389 (2003).
58. Sato, S. et al. TAK1 is indispensable for development of T cells and prevention
of colitis by the generation of regulatory T cells. Int. Immunol. 18, 1405–1411
59. wan, Y.Y., Chi, H., Xie, M., Schneider, M.D. & Flavell, r.A. The kinase TAK1 inte-
grates antigen and cytokine receptor signaling for T cell development, survival and
function. Nat. Immunol. 7, 851–858 (2006).
60. Haxhinasto, S., Mathis, D. & Benoist, C. The AKT-mTOr axis regulates de novo dif-
ferentiation of CD4+Foxp3+ cells. J. Exp. Med. 205, 565–574 (2008).
61. Sauer, S. et al. T cell receptor signaling controls Foxp3 expression via Pi3K, Akt, and
mTOr. Proc. Natl. Acad. Sci. USA 105, 7797–7802 (2008).
62. Zheng, X.X. et al. Favorably tipping the balance between cytopathic and regulatory
T cells to create transplantation tolerance. Immunity 19, 503–514 (2003).
63. Battaglia, M., Stabilini, A. & roncarolo, M.G. rapamycin selectively expands
CD4+CD25+FoxP3+ regulatory T cells. Blood 105, 4743–4748 (2005).
64. Strauss, L. et al. Selective survival of naturally occurring human CD4+CD25+Foxp3+
regulatory T cells cultured with rapamycin. J. Immunol. 178, 320–329 (2007).
65. Qu, Y. et al. The effect of immunosuppressive drug rapamycin on regulatory
CD4+CD25+Foxp3+ T cells in mice. Transpl. Immunol. 17, 153–161 (2007).
66. Delgoffe, G.M. mTor diferentially regulates effector and regulatory T cell lineage
commitment. Immunity (in the press).
67. Huehn, J., Polansky, J.K. & Hamann, A. epigenetic control of FOXP3 expression: the
key to a stable regulatory T-cell lineage? Nat. Rev. Immunol. 9, 83–89 (2009).
68. Hori, S. rethinking the molecular definition of regulatory T cells. Eur. J. Immunol.
38, 928–930 (2008).
69. Gavin, M.A. et al. Foxp3-dependent programme of regulatory T-cell differentiation.
Nature 445, 771–775 (2007).
70. Bacchetta, r. et al. Defective regulatory and effector T cell functions in patients with
FOXP3 mutations. J. Clin. Invest. 116, 1713–1722 (2006).
71. Haribhai, D. et al. A central role for induced regulatory T cells in tolerance induction
in experimental colitis. J. Immunol. 182, 3461–3468 (2009).
72. Allan, S.e., Song-Zhao, G.X., Abraham, T., McMurchy, A.N. & Levings, M.K. inducible
reprogramming of human T cells into Treg cells by a conditionally active form of
FOXP3. Eur. J. Immunol. 38, 3282–3289 (2008).
73. Davidson, T.S., Dipaolo, r.J., Andersson, J. & Shevach, e.M. Cutting edge: iL-2 is
essential for TGF-β-mediated induction of Foxp3+ T regulatory cells. J. Immunol.
178, 4022–4026 (2007).
74. Floess, S. et al. epigenetic control of the foxp3 locus in regulatory T cells. PLoS Biol.
5, e38 (2007).
75. Yu, A., Zhu, L., Altman, N.H. & Malek, T.r. A low interleukin-2 receptor signaling
threshold supports the development and homeostasis of T regulatory cells. Immunity
30, 204–217 (2009).
76. Chen, w. et al. Conversion of peripheral CD4+CD25– naive T cells to CD4+CD25+
regulatory T cells by TGF-β induction of transcription factor Foxp3. J. Exp. Med. 198,
77. Fantini, M.C. et al. Cutting edge: TGF-β induces a regulatory phenotype in CD4+CD25–
T cells through Foxp3 induction and down-regulation of Smad7. J. Immunol. 172,
78. Kretschmer, K. et al. inducing and expanding regulatory T cell populations by foreign
antigen. Nat. Immunol. 6, 1219–1227 (2005).
79. Apostolou, i. & von Boehmer, H. in vivo instruction of suppressor commitment in
naive T cells. J. Exp. Med. 199, 1401–1408 (2004).
80. verginis, P., McLaughlin, K.A., wucherpfennig, K.w., von, B.H. & Apostolou, i.
induction of antigen-specific regulatory T cells in wild-type mice: visualization and
targets of suppression. Proc. Natl. Acad. Sci. USA 105, 3479–3484 (2008).
81. Mucida, D. et al. Oral tolerance in the absence of naturally occurring Tregs. J. Clin.
Invest. 115, 1923–1933 (2005).
82. Sun, C.M. et al. Small intestine lamina propria dendritic cells promote de novo
generation of Foxp3 T reg cells via retinoic acid. J. Exp. Med. 204, 1775–1785
83. Curotto de Lafaille, M.A. et al. Adaptive Foxp3+ regulatory T cell-dependent and
-independent control of allergic inflammation. Immunity 29, 114–126 (2008).
84. Knoechel, B., Lohr, J., Kahn, e., Bluestone, J.A. & Abbas, A.K. Sequential devel-
opment of interleukin 2-dependent effector and regulatory T cells in response to
endogenous systemic antigen. J. Exp. Med. 202, 1375–1386 (2005).
85. Curotto de Lafaille, M.A., Lino, A.C., Kutchukhidze, N. & Lafaille, J.J. CD25– T cells
generate CD25+Foxp3+ regulatory T cells by peripheral expansion. J. Immunol. 173,
86. Singh, B. et al. Control of intestinal inflammation by regulatory T cells. Immunol.
Rev. 182, 190–200 (2001).
87. Fisson, S. et al. Continuous activation of autoreactive CD4+CD25+ regulatory T cells
in the steady state. J. Exp. Med. 198, 737–746 (2003).
88. Lathrop, S.K., Santacruz, N.A., Pham, D., Luo, J. & Hsieh, C.S. Antigen-specific
peripheral shaping of the natural regulatory T cell population. J. Exp. Med. 205,
89. wong, J., Mathis, D. & Benoist, C. TCr-based lineage tracing: no evidence for conver-
sion of conventional into regulatory T cells in response to a natural self-antigen in
pancreatic islets. J. Exp. Med. 204, 2039–2045 (2007).
90. Korn, T. et al. Myelin-specific regulatory T cells accumulate in the CNS but fail to
control autoimmune inflammation. Nat. Med. 13, 423–431 (2007).
91. Suffia, i., reckling, S.K., Salay, G. & Belkaid, Y. A role for CD103 in the retention
of CD4+CD25+ Treg and control of Leishmania major infection. J. Immunol. 174,
92. Taylor, M.D. et al. early recruitment of natural CD4+ Foxp3+ Treg cells by infective
larvae determines the outcome of filarial infection. Eur. J. Immunol. 39, 192–206
93. veldhoen, M., Hocking, r.J., Atkins, C.J., Locksley, r.M. & Stockinger, B. TGFβ in
the context of an inflammatory cytokine milieu supports de novo differentiation of
iL-17-producing T cells. Immunity 24, 179–189 (2006).
94. Bettelli, e. et al. reciprocal developmental pathways for the generation of pathogenic
effector TH17 and regulatory T cells. Nature 441, 235–238 (2006).
95. Zhou, L. et al. iL-6 programs TH-17 cell differentiation by promoting sequential
engagement of the iL-21 and iL-23 pathways. Nat. Immunol. 8, 967–974 (2007).
96. Yang, X.O. et al. Molecular antagonism and plasticity of regulatory and inflammatory
T cell programs. Immunity 29, 44–56 (2008).
97. Zhou, L. et al. TGF-β-induced Foxp3 inhibits TH17 cell differentiation by antagonizing
rOrγt function. Nature 453, 236–240 (2008).
98. Tsuji, M. et al. Preferential generation of follicular B helper T cells from Foxp3+ T
cells in gut Peyer’s patches. Science 323, 1488–1492 (2009).
99. Komatsu, N. et al. Heterogeneity of natural Foxp3+ T cells: a committed regulatory
T-cell lineage and an uncommitted minor population retaining plasticity. Proc. Natl.
Acad. Sci. USA 106, 1903–1908 (2009).
100. Beriou, G. et al. iL-17-producing human peripheral regulatory T cells retain suppres-
sive function. Blood 113, 4240–4249 (2009).
101. Lehmann, J. et al. expression of the integrin αeβ7 identifies unique subsets of CD25+
as well as CD25– regulatory T cells. Proc. Natl. Acad. Sci. USA 99, 13031–13036
102. Sather, B.D. et al. Altering the distribution of Foxp3+ regulatory T cells results in
tissue-specific inflammatory disease. J. Exp. Med. 204, 1335–1347 (2007).
103. Siewert, C. et al. experience-driven development: effector/memory-like αe+Foxp3+
regulatory T cells originate from both naive T cells and naturally occurring naive-like
regulatory T cells. J. Immunol. 180, 146–155 (2008).
104. Beyersdorf, N., Ding, X., Tietze, J.K. & Hanke, T. Characterization of mouse CD4
T cell subsets defined by expression of KLrG1. Eur. J. Immunol. 37, 3445–3454
105. Siegmund, K. et al. Migration matters: regulatory T-cell compartmentalization deter-
mines suppressive activity in vivo. Blood 106, 3097–3104 (2005).
106. Dudda, J.C., Perdue, N., Bachtanian, e. & Campbell, D.J. Foxp3+ regulatory T cells
maintain immune homeostasis in the skin. J. Exp. Med. 205, 1559–1565 (2008).
107. Zheng, Y. et al. regulatory T-cell suppressor program co-opts transcription factor
irF4 to control TH2 responses. Nature 458, 351–356 (2009).
108. Koch, M.A. et al. The transcription factor T-bet controls regulatory T cell homeostasis
and function during type 1 inflammation. Nat. Immunol. 10, 595–602 (2009).
109. Glimcher, L.H. Trawling for treasure: tales of T-bet. Nat. Immunol. 8, 448–450
110. Busslinger, M. Transcriptional control of early B cell development. Annu. Rev.
Immunol. 22, 55–79 (2004).
111. Schneider, M.A., Meingassner, J.G., Lipp, M., Moore, H.D. & rot, A. CCr7 is required
for the in vivo function of CD4+CD25+ regulatory T cells. J. Exp. Med. 204, 735–745
112. Peng, S.L., Szabo, S.J. & Glimcher, L.H. T-bet regulates igG class switching and
pathogenic autoantibody production. Proc. Natl. Acad. Sci. USA 99, 5545–5550
nature immunology volume 10 number 7 july 2009
© 2009 Nature America, Inc. All rights reserved.