f o C u s o n To l e r A n C e i n T r A n s p l A n TAT i o n

An update on regulatory T cells in transplant

tolerance and rejection
Xian Chang Li and Laurence A. Turka
Abstract | Several types of T cells with immunosuppressive properties have been identified, but FOXP3+
regulatory T (TREG) cells have emerged as a dominant cell type; they are critically involved in the induction
and maintenance of immune tolerance. Manipulation of this cell type for the induction of transplant
tolerance including renal transplant tolerance has attracted considerable attention. Studies in this area have
demonstrated unexpected complexities, and attempts to translate TREG cells towards clinical utility have met
with unanticipated difficulties. In this Review, a broad overview is provided on recent progress in the study
of TREG cells, focusing on challenges, opportunities, and emerging approaches in exploiting T REG cells for the
induction of transplant tolerance.
Li, X. C. & Turka, L. A. Nat. Rev. Nephrol. 6, 577–583 (2010); published online 3 August 2010; doi:10.1038/nrneph.2010.101

Introduction
one of the major goals in transplantation is the induction challenges in targeting treG cells for the induction of
of tolerance, that is, indefinite allograft survival without transplant tolerance.
the need for life-long immunosuppression. over the although several subsets of t cells have been identi-
past 10 years, the approach of the scientific community fied to date that have immune regulatory properties, in
towards achieving this goal has evolved from one that this review focus will be placed specifically on FoXP3+
focused on inhibition of t-effector cells to one that also treG cells. this FoXP3+ regulatory cell type is certainly
involves promotion of regulatory t (treG) cells.1 the the most studied and best understood, and the existence
current paradigm is that the balance of alloantigen- of an autoimmune syndrome called immune dysregula-
reactive treG cells and t-effector cells ultimately deter- tion, polyendocrinopathy, enteropathy, X-linked (iPeX)
mines whether the graft is accepted or rejected, and that syndrome in humans who lack a functional FOXP3 gene
promoting the development and maintenance of regula- is a clear evidence of its role in peripheral tolerance and
tory responses represents a promising approach for the immune homeostasis.5 in this review we provide an
induction of transplantation tolerance. overview of treG cells and tolerance induction in various
although the existence of treG cells is indisputable, transplant models including kidney transplantation. the
harnessing them for therapeutic purposes has not reason why discussion does not solely focus on kidney
been straightforward. For example, the local micro- transplantation is because progress and issues in other
environment in which treG cells reside can have models are applicable to renal transplantation and vice
considerable influence on their functional status. Certain versa. this review highlights recent progress in the
inflammatory cytokines can either disarm treG cells study of FoXP3+ treG cells, and summarizes emerging
or render t-effector cells resistant to suppression. 2 approaches for their manipulation in the induction of
additionally, treG cells are extremely heterogeneous transplant tolerance.
and consist of many different subsets. moreover, some of
these subsets exhibit considerable plasticity in that they T cells with regulatory properties
are neither terminally differentiated nor hardwired for although multiple types of immune regulatory t cells
suppression;3 their regulatory programs can be readily exist, they are often broadly divided into two categories—
turned off or reprogrammed.4 as a consequence, former FoXP3– and FoXP3+ regulatory cells. FoXP3– regulatory
treG cells can, under some circumstances, such as in an cells include CD4+il-10-producing tr1 cells,6 a subset
The Transplant
inflammatory milieu, acquire the phenotype and func- of CD8+ t cells, and CD3+CD4–CD8– double-negative Institute, Beth Israel
tional capabilities of t-effector cells and mediate tissue t cells.7 these cells suppress t-cell activation through Deaconess Medical
Center, Harvard
damage. Clearly, these complexities present unexpected a variety of different mechanisms and have been shown Medical School,
to contribute to transplant tolerance in selected animal 330 Brookline Avenue,
models. in addition, natural killer t cells that are also CLS‑605, Boston,
MA 02215, USA
Competing interests FoXP3– can have potent immune suppressive activities (X. C. li, l. A. Turka).
L. A. Turka declares associations with the following companies:
in some transplant models.8 although the FoXP3 protein
Biogen IDEC, Bristol–Myers Squibb, GlaxoSmithKline and Correspondence to:
Novartis. See the article online for full details of the can be found in both CD4+ and CD8+ t cells, the former X. C. Li
relationships. X. C. Li declares no competing interests. (that is, CD4+FoXP3+ cells, referred to as treG cells) [email protected]

nature reviews | nephrology volume 6 | oCtoBer 2010 | 577

© 2010 Macmillan Publishers Limited. All rights reserved
reVieWs

Key points adaptive treG cells, or more commonly, induced (i)treG

■ Regulatory T (TREG) cells are indispensable in transplant tolerance
cells) (Figure 1).12 ntreG cells and itreG cells share some
common features in terms of phenotype and function.
■ TREG cells are heterogeneous and are not inherently stable; their suppressive
programs can be turned off or they may be reprogrammed to become T‑effector
For example, both types of cells express FoXP3 and both
cells potently suppress the proliferation of t-effector cells.12
■ Some effector cell types, such as memory T cells, can evade TREG‑cell‑mediated
additionally, t-cell receptor (tCr) signaling is required
suppression for the induction of both ntreG cells and itreG cells, as
targeted mutations that disrupt proximal tCr signal-
■ Mechanisms that regulate the in vivo induction and stability of TREG cells are
poorly defined ing events often result in a complete absence of thymic
Foxp3+ t cells.13 similarly, suppression of tCr signal-
■ The mechanism by which FOXP3+ TREG cells interact with other types of
regulatory cells in transplant tolerance is unknown ing by calcineurin inhibitors blocks the conversion of
t-effector cells to itreG cells.14 Finally, cytokines such as
■ Potential complications of immunodeficiency exist from nonspecific expansion
of TREG cells
tGF-β and il-2 are critical to intrathymic development
of ntreG cells15 and are also mandatory for the induc-
tion of itreG cells in the periphery.16 this observation
is consistent with the finding that the FOXP3 promoter
region contains binding sites for smaD proteins and
stat5, key molecules involved in tGF-β and il-2
FOXP3+ TREG cells signaling, respectively.
FOXP3, CD25, GITR,
CTLA-4, IL-10, CD39,
OX40, galectin-9 Key features of nTreg and iTreg cells
ntreG cells and itreG cells also have striking differ-
ences (Figure 1). ntreG cells are selected and matured
in the thymus by autoantigens, before being exported to
the periphery. thus, the tCr repertoire of these cells is
nTREG cells iTREG cells
primarily autoreactive.17 By contrast, as itreG cells are
■ Thymus-derived ■ Generated in the periphery converted from conventional t-effector cells in the
■ Selected by autoantigens ■ Derived from T-effector cells periphery, they supposedly have a similar tCr repertoire
■ Possess an autoreactive TCR repertoire ■ Induction requires TCR, CD28 signaling,
■ Cross-reactive to alloantigens and certain cytokines (e.g. TGF-β, IL-2) to that of the t-effector cells from which they are derived.
■ Suppress autoimmunity ■ Positively and negatively regulated by many therefore, although itreG cells are initially selected by
■ Contact-dependent fashion in suppression other pathways
autoantigens in the thymus, as are all t cells, they may be
Figure 1 | Key features and differences between nTREG cells and iTREG cells. Both considered to be more skewed toward foreign antigens.
subsets express the transcription factor FOXP3 and a variety of other cell surface this difference in tCr specificity is important in trans-
factors, but they are strikingly different in origin, induction, TCR repertoire and in plantation where tolerance is often defined by donor
antigen specificity. Abbreviations: iTREG cell, induced regulatory T cell; nTREG cell, antigen specificity.18 moreover, many other factors and
natural regulatory T cell; TREG cell, regulatory T cell; TCR, T‑cell receptor.
pathways facilitate the induction of itreG cells, includ-
ing retinoic acid,19 leukemia inhibitory factor (liF),20
have emerged as the dominant focus of research and Ctla-4, and PD-1–PD-l1 interaction.21 equally impor-
application in both preclinical and clinical studies.9 tant is the finding that induction of itreG cells can be
inhibited by several mechanisms, including by certain
CD4+foXp3+ Treg cells cytokines (for example, il-4, il-6, iFn-γ), by oX40 and
the discovery that the transcription factor FoXP3 was t cell-ig mucin molecule-1 (tim-1) costimulation or by
both necessary and sufficient for treG-cell function in an sustained akt–mtor activation.4,22,23 intriguingly, these
entire class of CD4+ t cells, firmly re-established the study pathways seem to have little effect on the intrathymic
of immune regulation in general, and sharply focused development of ntreG cells. in addition, knockout of
the spotlight on these cells in particular.10 Phenotypically, the gene that encodes the membrane-associated guany-
these treG cells express several cell surface markers such late kinase Carma‑1 results in selective deficiency of
as CD25, Ctla-4, Gitr (glucocorticoid-induced tumor ntreG cells in the thymus, although CD4 + Foxp3 –
necrosis factor receptor), CD134, CD103, CD39 and t-effector cells in the periphery can be readily converted
CD73, many of which are also expressed by activated to Foxp3+ itreG cells by tGF-β,13 suggesting key differ-
CD4+ t-effector cells, and are therefore not treG-cell ences in the development of ntreG cells and itreG cells.
specific. under basal conditions, treG cells account for interestingly, itreG cells are particularly unstable and tend
approximately 5–10% of the total CD4+ t cells in the to lose FoXP3 expression more easily than ntreG cells,
periphery, and can expand via lymphopenia-driven or which is probably related to epigenetic differences in
antigen-driven proliferation in vivo.11 the FOXP3 gene in both subsets.24 such differences may
even as defined by universal FoXP3 expression, be important in the selective manipulation of ntreG cells
treG cells are not a uniform population, but are broadly and itreG cells by therapeutic means.
subdivided into those that developed in the thymus with few exceptions, conversion of t-effector cells to
(so-called natural (n)treG cells) and those that were itreG cells in vivo may be surprisingly rare.25 Despite the
converted from t-effector cells in the periphery (termed fact that t-effector cells are capable of becoming itreG

578 | OCTOBER 2010 | vOlumE 6 www.nature.com/nrneph

© 2010 Macmillan Publishers Limited. All rights reserved
 f o C u s o n To l e r A n C e i n T r A n s p l A n TAT i o n

cells under certain defined conditions,16 in vivo studies amounts of immune suppressive cytokines, including
showing consistent and reproducible induction of itreG tGF-β, il-10, and il-35, which are known to inhibit a
cells, either in naïve hosts or under tolerizing conditions, wide spectrum of cellular activities.31 tGF-β exerts broad
are few and far between. tolerizing protocols are hoped antiproliferative and anti-inflammatory effects. il-10
to induce prominent itreG cells that confer donor spe- strongly inhibits activation of macrophages and dendritic
cific tolerance in transplant models, but clear-cut evi- cells, and il-35 is a key mediator of treG-cell-induced
dence supporting this claim in tolerant recipients, is still immunosuppression.32 Despite advances in our under-
lacking. in fact, a lack of information about the in vivo standing of treG-cell function, the processes by which
conditions that promote the conversion of t-effector cells this catalog of in vitro mechanisms contributes to in vivo
to treG cells in transplant recipients indicates that more immunosuppression by treG cells is still not known.
studies are warranted in this area. in this regard, it should
be noted that the relative contribution of ntreG cells and Harnessing TREG cells for tolerance
itreG cells to the peripheral pool of treG cells and the Despite what many immunologists believe to be the tre-
relevance of each subset to transplant tolerance are not mendous potential of treG cells for inducing transplant
known. whether ntreG cells and itreG cells are syner- tolerance, translating this strategy into reality has not
gistic in promoting tolerance or are merely redundant been straightforward. many challenging issues face this
is not yet clear; however this important issue requires area that require further study.
further investigation.
nTreg cells versus iTreg cells
Mechanisms of TREG-cell action when compared in a variety of assays in vitro and in vivo,
Both innate and adaptive immune cells are targets ntreG cells and itreG cells show comparable potency in
of treG-cell-mediated suppression, and treG cells are suppressing t-effector cells.4 However, as noted above,
known to employ a variety of mechanisms to mediate the relative importance of ntreG cells and itreG cells in
these effects. treG cells directly suppress many functions transplant models remains unknown. whether both
of CD4+ and CD8+ t cells, ranging from their prolifera- subsets are required or whether one subset predomi-
tion to their differentiation into t-helper (tH)-1, tH2, and nates the other in the induction of transplant tolerance
tH17 subsets. in some cases, treG cells induce apoptosis of is not clear. a subset of ntreG cells can be alloreactive,
responding t cells.26 treG cells also suppress the activation presumably owing to their cross-reactivity to allo-
of B cells, thereby inhibiting humoral immune responses. antigens. in fact, in several transplant models, global
other prominent targets of treG cells include dendritic depletion of ntreG cells prevents tolerance induction to
cells, macrophages, natural killer cells, mast cells, and allografts.9 the exact proportion and identity of ntreG
osteoblasts.26 treG cells are also involved in tissue repair cells that are alloreactive is unclear, however, as no
and in the resolution of tissue inflammation,27 suggesting specific markers exist to distinguish alloreactive ntreG
a potential role for treG cells in the regulation of non- cells from nonalloreactive ones. ideally, therapies that
immune cells. in fact, treG cells can inhibit the develop- promote treG-cell-mediated tolerance should selec-
ment of transplant vasculopathy,28 a complex process that tively and specifically activate and expand alloreactive
involves many immune and nonimmune cells, supporting ntreG cells. targeting ntreG cells en-mass in transplant
a broad role for treG cells in tissue remodeling. models, though beneficial, may also induce unwanted
the mechanism by which treG cells regulate such effects owing to a stimulation of nonalloreactive treG
diverse cell types both inside and outside the immune cells that could potentially render patients susceptible to
system is an area of considerable interest. treG cells prob- immunodeficient complications.
ably employ several different mechanisms to suppress on the other hand, itreG cells are derived from
pathogenic t-effector cells. In vitro assays have demon- t-effector cells and should have the same antigen speci-
strated that activation of treG cells via tCr stimulation ficity as alloreactive t-effector cells. thus, itreG cells
is required to mediate suppression of t-effector cells and may have a particularly important role in donor-specific
that this suppression requires strict cell–cell contact.26 tolerance. in fact, studies that have tried to expand itreG
under some conditions, treG cells have been shown to cells ex vivo and then adoptively transfer them to induce
deprive t-effector cells of survival and growth factors or tolerance to allografts in vivo have considerable attrac-
to directly kill activated t cells via granzyme-dependent tion, as in selected models itreG cells can be key players
mechanisms. Furthermore, treG cells express CD39 and in tolerance induction.33 However, the paucity of itreG
CD73, the ectoenzymes that break down the extracellular cells in vivo, even under tolerizing conditions in trans-
atP into adenosine.29 this process has been shown to plant models, raises concerns. one issue is that itreG
turn an atP-rich inflammatory milieu to one that is cells must compete with ntreG cells in vivo for space,
immunosuppressive, as adenosine inhibits the activa- resources, and survival, and in this setting, itreG cells
tion of dendritic cells and macrophages, which in turn may have a competitive disadvantage. another concern
prevents t-cell priming. treG cells also express Ctla-4 is that the induction of itreG cells can be diverted, leading
on their surface, which can directly engage the periph- to the formation of pathogenic tH17 cells in the presence
eral membrane protein B7 on antigen-presenting cells of inflammatory cytokines.34 whether blocking tissue
(aPCs) to inhibit aPC activation via different mech- inflammation can enhance induction of itreG cells in vivo
anisms.30 in addition, treG cells can produce copious remains to be studied.

nature reviews | nephrology volume 6 | oCtoBer 2010 | 579

© 2010 Macmillan Publishers Limited. All rights reserved
reVieWs

FOXP3, IL-2 and colleagues used an elegant lineage tracking tech-

miRNA
Epigenetics nique to demonstrate that a notable fraction of ntreG
cells fail to maintain Foxp3 expression; such ‘ex-treG cells’
produced substantial amounts of iFn-γ and mediated
the development of autoimmune diabetes.35 the treG-cell
pool therefore seems to contain a population of com-
Immune suppression mitted treG cells and a small fraction of uncommitted
treG cells. such uncommitted treG cells seem to easily
lose FoXP3 expression and become t-effector cells
(Figure 2). whether the a priori identification of uncom-
mitted treG cells is possible, is not known, and although
demethylation of the FOXP3 locus may prove to be a
useful marker to identify uncommitted treG cells, its
Cytopathic effects assessment in live cells is difficult.24 although all of the
factors involved in the regulation of ntreG-cell stability
are probably not known, certain inflammatory cytokines
(for example, il-1, il-6, tnF) and costimulatory mol-
ecules (for example, oX40, tim-1) seem to be critically
involved.40 whether treG cells can be stabilized in vivo by
blocking these pathways remains to be determined.
Further concerns are held with regard to the instabil-
ity of itreG cells, as a genome-wide analysis of itreG cells
Unstable TREG cells
and ntreG cells identified many epigenetic differences
critical to treG-cell stability.24 For example, CpG islands,
Committed TREG cells histone methylation and acetylation as well as mirna
status are major mechanisms that contribute to the stabil-
Figure 2 | Lineage plasticity of TREG cells in the periphery. Multiple pathways and ity of FoXP3+ treG cells,41 whereas itreG cells exhibit an
mechanisms are involved in regulating the peripheral pool of TREG cells, including
epigenetic signature similar to that of t-effector cells
miRNA, epigenetic mechanisms, and pathways that involve FOXP3 and IL‑2. The
TREG cell pool consists of committed TREG cells that are relatively stable and
as opposed to that of treG cells.24 this variation may
dedicated to suppression and unstable TREG cells. Unstable TREG cells can lose be one of the key reasons for the instability of itreG
FOXP3 and acquire T‑effector cell functions. Abbreviations: miRNA, microRNA; cells, and, therefore, the paucity of itreG cells in vivo
TREG cell, regulatory T cell. in transplant recipients, even under tolerizing condi-
tions. Determining whether such mechanisms can be
targeted to stabilize itreG-cell populations once they are
lineage plasticity of Treg cells induced in transplant models to promote tolerance is of
as alluded to above, although treG cells were initially critical importance.
believed to be highly specialized and fully committed to
suppression of t-effector cells at all times, in fact their presentation of target alloantigen to Treg cells
lineage and suppressive programs exhibit unexpected a key issue unique to transplantation is the presence
plasticity.3 For example, ntreG cells can lose Foxp3 expres- of donor and host aPCs in transplant recipients; they
sion, with a resultant loss of immunosuppressive function, process and present alloantigens in strikingly differ-
and a potential acquisition of effector function, rendering ent fashions. 42 Donor aPCs use the direct pathway
them capable of causing tissue destruction.35 another sur- whereas host aPCs use the indirect pathway to stimulate
prise is that expression of Foxp3, though instrumental to t-effector cells and treG cells. this difference in antigen
the development of treG cells, is not an on-and-off switch presentation may have considerable influence on the
to treG-cell functionality. in fact, differences in levels of induction, homeostasis, and function of treG cells.43
Foxp3 expression also profoundly influence their sup- whether donor aPCs and host aPCs are equally potent
pressor programs, and reduced Foxp3 expression tends in the induction of itreG cells or in the expansion of ntreG
to induce impaired treG-cell functions.36 cells remains to be defined, nor is it clear whether donor
a number of studies exist demonstrating functional and host aPCs differentially affect ntreG cells and itreG
instability of ntreG cells. stimulation of highly purified cells. in some models, treG cells that induce transplant
ntreG cells in vitro in the presence of il-6 resulted in the tolerance demonstrate indirect antigen specificity, 44
loss of Foxp3 and such treG cells then became il-17- implying that host aPCs that present allopeptides indi-
producing t-effector cells.37 moreover, genetic muta- rectly may be more important than those that present
tions that disrupt the micro (mi)rna pathways cause allopeptides directly in this process. aPCs may be key
treG cells to lose their suppressive function despite main- targets of treG-cell-mediated suppression in vivo, yet it
tenance of Foxp3 expression.38 in humans, CD4+CD25low is not known if treG cells are equally effective in suppres-
cells that express lower levels of FoXP3 than normal sion of donor and host aPCs, nor whether ntreG cells and
CD4+CD25+ cells can readily lose FoXP3 expression itreG cells recognize the same or different alloantigens in
and acquire il-17-producing ability.39 in 2009, Bluestone the induction of tolerance in transplant models.

580 | OCTOBER 2010 | vOlumE 6 www.nature.com/nrneph

© 2010 Macmillan Publishers Limited. All rights reserved
 f o C u s o n To l e r A n C e i n T r A n s p l A n TAT i o n

Induction Tolerance
TREG pool

nTREG cells

nTREG cells
Alloantigen specific
Stable and long lasting
Naïve T cells

Reduction in number
of T-effector cells

T-effector cells
iTREG cells

Time
Figure 3 | Evolution of TREG cells in transplant tolerance. Induction of donor‑specific TREG cells is critically important in
tolerance, and both nTREG cells that cross‑react to alloantigens and iTREG cells that are derived from alloreactive T‑effector
cells may contribute to the alloreactive TREG‑cell pool. The longevity and alloantigen specificity of such TREG cells is essential
to tolerance maintenance. Abbreviations: iTREG cell, induced regulatory T cell; nTREG cell, natural regulatory T cell; TREG cell,
regulatory T cell.

Treg cells within and outside of lymphoid tissue to be long-lived, there is, in fact, ongoing slow but sub-
in addition to being present in lymphoid tissues, treG stantial attrition in the periphery.48 little is known about
cells are also found in large numbers at extralymphoid the lifespan of alloreactive treG cells, but they are unlikely
sites, although the function of extra-lymphoid treG to defy the factors that govern the life and death of t cells.
cells is only beginning to be studied. treG cells at non- one potential mechanism that could overcome this
lymphoid sites are not dormant cells. For example, treG limitation is ‘infectious tolerance’ where treG cells per-
cells in adipose tissue have been shown to regulate met- petuate tolerance by recruiting a new cohort of t cells to
abolic processes;45 such treG cells are different from those become alloreactive treG cells to constantly replenish
in the spleen. interestingly, treG cells in fat from obese the treG-cell pool, so that tolerance can be continuously
mice are altered in both phenotype and function.45 in the enforced in vivo.49 in selected models, this mechanism
transplant setting, treG cells can be found residing in is clearly potent in the maintenance of a tolerant state.50
tolerant grafts,46 but the relative importance of treG cells whether the newly recruited treG cells are itreG cells,
in the grafts versus those in the lymphoid tissues is not (that is, recruited from naïve t-effector cells) or are
known. whether they are the same type of cells present at derived from the ntreG-cell pool, or a combination of
different sites or they are inherently different is not clear. both, is not known (Figure 3). Clearly, more studies in
it is also not apparent if they are functionally equivalent. this area are undoubtedly needed.
in most models of immune responses, priming of t cells
for rejection takes place in the draining lymph nodes, and resistance to Treg-cell-mediated suppression
treG cells at these sites may be particularly important for ample exceptions exist where competent treG cells fail
the suppression of t-effector cells. treG cells may then to restrain t-effector cells. For example, treG cells often
migrate to grafts where they have a key role in the control completely fail to inhibit the responses of memory
of intragraft inflammation, which perpetuates the induc- t cells.51 Furthermore, t-effector cells activated under
tion of new treG cells in situ. Graft acceptance has been inflammatory conditions are highly resistant to treG-cell-
suggested to require the presence of intragraft treG cells. mediated immunosuppression,52 suggesting that target
a 2009 study by Bromberg and colleagues demonstrated a cells can overturn the suppressive function of treG cells.
sequential migration of treG cells from blood to the trans- t-effector cells can, therefore, ‘fight back’ to evade treG
planted grafts and then to the draining lymph nodes in cells. in addition, certain cells of the innate immune
order to achieve tolerance,47 which suggests that the role system, for example, natural killer cells, can kill treG
of treG cells in transplant settings may be very dynamic. cells upon activation.53 under such conditions, trans-
plant tolerance is difficult to establish, and alternative
longevity and homeostasis of Treg cells in vivo approaches that target cells that are resistant to treG-cell-
t cells, regardless of their phenotypes and functions, have mediated suppression or that restore their sensitivity to
a finite lifespan. even for memory t cells that are thought treG cells are important in the induction of tolerance.

nature reviews | nephrology volume 6 | oCtoBer 2010 | 581

© 2010 Macmillan Publishers Limited. All rights reserved
reVieWs

Differences between humans and mice and stability of expanded treG cells in vivo following
many key differences exist in treG cells between humans adoptive cell transfer. as discussed above, both can be
and mice. one of the outstanding features is that in technically challenging. nonetheless, this approach
humans, FoXP3 is a t-cell activation molecule rather undoubtedly holds great promise in harnessing treG
than an exclusive marker of treG cells, as stimulation of cells for therapeutic purposes. the second approach
highly purified human t-effector cells is associated with is to selectively and specifically stimulate treG cells
transient expression of FoXP3, and t cells that temporar- in vivo, by taking advantage of fundamental differences
ily acquire FoXP3 do not always show suppressive func- between the biology of treG cells and t-effector cells.
tions.54 evidence exists that human CD4+ treG cells and For example, treG cells constitutively express the high
non-treG cells can be segregated by differential expression affinity il-2 receptor, and therefore have a competitive
of CD25 and CD127 (the alpha chain of il-7 receptor) advantage in the utilization of il-2.59 in an islet trans-
where functional treG cells are confined exclusively to plant model in the mouse, an il-2–ig fusion protein or
the CD25+CD127– fraction.55,56 studies from sakaguchi’s an il-2–anti-il-2 complex that has a long half-life in vivo
research group have demonstrated additional strik- can promote robust expansion of treG cells, which are
ing differences between human and mouse treG cells.57 permissive for islet allograft tolerance.60,61 although in
Based on levels of FoXP3 expression and various CD45 theory, il-2 might also stimulate t-effector cells, studies
isoforms, they have identified at least three different of il-2 in humans suggest a preferential ability to expand
human treG-cell subsets. in essence, resting treG cells are treG cells. this effect may be even more prominent when
CD45ra+FoXP3low, but they readily acquire a CD45ra– il-2 is combined with mtor inhibitors such as rapa-
FoXP3 high phenotype upon activation; both subsets mycin. 61 in addition, other approaches that include
are potent suppressor cells when deliberately tested. inhibition of innate immune cells, modification of epi-
interestingly, the activated CD45ra–FoXP3high treG cells genetic mechanisms, and prevention of memory t-cell
are extremely sensitive to apoptotic cell death. However, activation are also being actively explored and likely to
the CD45ro+FoXP3low subset does not have suppressive be important facets of promoting treG-cell-mediated
activities, but instead produces proinflammatory cyto- tolerance in transplantation.
kines, and therefore exhibits effector functions.57 Clearly,
these unique features should be carefully considered in Conclusions
targeting human treG cells in the transplant setting. also, treG cells are a key cell type in the induction and main-
different isoforms of FoXP3 exist in humans but not in tenance of immune tolerance, thus modulation of treG
mice,58 suggesting that FoXP3 and the FoXP3-mediated cells could provide new strategies in creating trans-
programs may be regulated differently in humans and in plant tolerance. many issues and challenges still need
mice. these differences inject a key note of caution to be addressed in this area, however. as such barriers
in extrapolating strategies and successes from mouse continue to be identified and overcome, therapeutic
models into approaches in the clinic. approaches harnessing the power of treG cells may even-
tually become a key element of a multifaceted strategy in
Emerging approaches the induction of transplant tolerance.
two broad approaches exist in the use of treG cells
to promote transplant tolerance. the first approach is to
Review criteria
first expand treG cells in vitro and then apply expanded
treG cells as a cell therapy in vivo. the advantage of this The information discussed in this Review is based on
approach is that antigen-specific treG cells can be created the experience of the authors and literature accumulated
in vitro using donor antigens. this approach, however, over their years working in this field. No specific database
searches were performed.
relies on the successful cultivation of treG cells in vitro

1. Lechler, R., Garden, O. A. & Turka, L. A. The 7. Zhang, Z. X., Yang, L., Young, K. J., DuTemple, B. disease by IL‑2 neutralization. J. Exp. Med. 201,
complementary roles of deletion and regulation & Zhang, L. Identification of a previously 723–735 (2005).
in transplantation tolerance. Nat. Rev. Immunol. unknown antigen‑specific regulatory T cells and 12. Curotto de Lafaille, M. A. & Lafaille, J. J. Natural
3, 147–158 (2003). its mechanism of suppression. Nat. Med. 6, and adaptive foxp3+ regulatory T cells: more of
2. Strom, T. B. & Koulmanda, M. Recently 782–789 (2000). the same or a division of labor? Immunity 30,
discovered T cell subsets cannot keep their 8. Seino, K.‑I. et al. Requirement for natural killer 626–635 (2009).
commitments. J. Am. Soc. Nephrol. 20, T (NKT) cells in the induction of allograft 13. Molinero, L. L. et al. CARMA1 controls an early
1677–1680 (2009). tolerance. Proc. Natl Acad. Sci. USA 98, checkpoint in the thymic development of FoxP3+
3. Zhou, X., Bailey‑Bucktrout, S., Jeker, L. T. & 2577–2581 (2001). regulatory T cells. J. Immunol. 182, 6736–6743
Bluestone, J. A. Plasticity of CD4(+)Foxp3(+) 9. Wood, K. J. & Sakaguchi, S. Regulatory T cells in (2009).
T cells. Curr. Opin. Immunol. 21, 281–285 (2009). transplantation tolerance. Nat. Rev. Immunol. 3, 14. Gao, W. et al. Contrasting effects of cyclosporine
4. Vu, M. D. et al. OX40 costimulation turns off 199–210 (2003). and rapamycin on de novo generation of antigen‑
Foxp3+ Tregs. Blood 110, 2501–2510 (2007). 10. Hori, S., Nomura, T. & Sakaguchi, S. Control of specific regulatory T cells. Am. J. Transplant. 7,
5. Sakaguchi, S. Naturally arising CD4+ regulatory regulatory T cell development by the 1722–1732 (2007).
T cells for immunologic self‑tolerance and transcription factor Foxp3. Science 299, 15. Liu, Y. et al. A critical function for TGF‑beta
negative control of immune responses. Annu. 1057–1061 (2003). signaling in the development of natural
Rev. Immunol. 22, 531–562 (2004). 11. Setoguchi, R., Hori, S., Takahashi, T. & CD4+CD25+Foxp3+ regulatory T cells. Nat.
6. Groux, H. et al. A CD4+ T‑cell subset inhibits Sakaguchi, S. Homeostatic maintenance of Immunol. 9, 632–640 (2008).
antigen‑specific T‑cell responses and prevents natural Foxp3+ CD25+ CD4+ regulatory T cells by 16. Chen, W. et al. Conversion of peripheral
colitis. Nature 389, 737–742 (1997). interleukin (IL)‑2 and induction of autoimmune CD4+CD25‑ naive T cells to CD4+CD25+

582 | OCTOBER 2010 | vOlumE 6 www.nature.com/nrneph

© 2010 Macmillan Publishers Limited. All rights reserved
 f o C u s o n To l e r A n C e i n T r A n s p l A n TAT i o n
regulatory T cells by TGF‑β induction of 33. You, S. et al. Adaptive TGF‑beta‑dependent 49. Qin, S. et al. “Infectious” transplantation
transcription factor Foxp3. J. Exp. Med. 198, regulatory T cells control autoimmune diabetes tolerance. Science 259, 974–977 (1993).
1875–1886 (2003). and are a privileged target of anti‑CD3 antibody 50. Honey, K., Cobbold, S. P. & Waldmann, H. CD40
17. Fontenot, J. D. et al. Regulatory T cell lineage treatment. Proc. Natl Acad. Sci. USA 104, ligand blockade induces CD4+ T cell tolerance
specification by the forkhead transcription factor 6335–6340 (2007). and linked suppression. J. Immunol. 163,
foxp3. Immunity 22, 329–341 (2005). 34. Veldhoen, M., Hocking, R. J., Atkins, C. J., 4805–4810 (1999).
18. Waldmann, H. Transplantation tolerance—where Locksley, R. M. & Stockinger, B. TGF‑beta in the 51. Yang, J. et al. Allograft rejection mediated by
do we stand? Nat. Med. 5, 1245–1248 (1999). context of an inflammatory cytokine milieu memory T cells is resistant to regulation. Proc.
19. Benson, M. J., Pino‑Lagos, K., Rosemblatt, M. & supports de novo differentiation of IL‑17 Natl Acad. Sci. USA 104, 19954–19959 (2007).
Noelle, R. J. All‑trans retinoic acid mediates producing cells. Immunity 24, 179–189 (2006). 52. Korn, T. et al. Myelin‑specific regulatory T cells
enhanced T reg cell growth, differentiation, and 35. Zhou, X. et al. Instability of the transcription accumulate in the CNS but fail to control
gut homing in the face of high levels of co‑ factor Foxp3 leads to the generation of autoimmune inflammation. Nat. Med. 13,
stimulation. J. Exp. Med. 204, 1765–1774 pathogenic memory T cells in vivo. Nat. 423–431 (2007).
(2007). Immunol. 10, 1000–1007 (2009). 53. Roy, S. et al. NK cells lyse T regulatory cells
20. Gao, W. et al. Treg versus Th17 lymphocyte 36. Wan, Y. Y. & Flavell, R. A. Regulatory T‑cell that expand in response to an intracellular
lineages are cross‑regulated by LIF versus IL‑6. functions are subverted and converted owing to pathogen. J. Immunol. 180, 1729–1736
Cell Cycle 8, 1444–1450 (2009). attenuated Foxp3 expression. Nature 445, (2008).
21. Francisco, L. M. et al. PD‑L1 regulates the 766–770 (2007). 54. Allan, S. E. et al. Activation‑induced FOXP3 in
development, maintenance, and function of 37. Yang, X. O. et al. Molecular antagonism and human T effector cells does not suppress
induced regulatory T cells. J. Exp. Med. 206, plasticity of regulatory and inflammatory T cell proliferation or cytokine production. Int.
3015–3029 (2009). programs. Immunity 29, 44–56 (2008). Immunol. 19, 345–354 (2007).
22. Wei, J. et al. Antagonistic nature of T helper 1/2 38. Zhou, X. et al. Selective miRNA disruption in 55. Seddiki, N. et al. Expression of interleukin (IL)‑2
developmental programs in opposing peripheral Treg cells leads to uncontrolled autoimmunity. and IL‑7 receptors discriminates between
induction of Foxp3+ regulatory T cells. Proc. Natl J. Exp. Med. 205, 1983–1991 (2008). human regulatory and activated T cells. J. Exp.
Acad. Sci. USA 104, 18169–18174 (2007). 39. Radhakrishnan, S. et al. Reprogrammed FoxP3+ Med. 203, 1693–1700 (2006).
23. Haxhinasto, S., Mathis, D. & Benoist, C. The T regulatory cells become IL‑17+ antigen‑ 56. Liu, W. et al. CD127 expression inversely
AKT‑mTOR axis regulates de novo differentiation specific autoimmune effectors in vitro and correlates with FoxP3 and suppressive function
of CD4+Foxp3+ cells. J. Exp. Med. 205, 565–574 in vivo. J. Immunol. 181, 3137–3147 (2008). of human CD4+ T reg cells. J. Exp. Med. 203,
(2008). 40. Li, X. C., Rothstein, D. M. & Sayegh, M. H. 1701–1711 (2006).
24. Lal, G. & Bromberg, J. S. Epigenetic mechanisms Costimulatory pathways in transplantation: 57. Miyara, M. et al. Functional delineation and
of regulation of Foxp3 expression. Blood 114, challenges and new developments. Immunol. differentiation dynamics of human CD4+ T cells
3727–3735 (2009). Rev. 229, 271–293 (2009). expressing the FoxP3 transcription factor.
25. Wong, J., Mathis, D. & Benoist, C. TCR‑based 41. Lu, L. F. et al. Foxp3‑dependent microRNA155 Immunity 30, 899–911 (2009).
lineage tracing: no evidence for conversion of confers competitive fitness to regulatory T cells 58. Allan, S. E. et al. The role of 2 FOXP3 isoforms
conventional into regulatory T cells in response by targeting SOCS1 protein. Immunity 30, in the generation of human CD4+ Tregs. J. Clin.
to a natural self‑antigen in pancreatic islets. 80–91 (2009). Invest. 115, 3276–3284 (2005).
J. Exp. Med. 204, 2039–2045 (2007). 42. Auchincloss, H. Jr & Sultan, H. Antigen 59. Bayer, A. L., Yu, A., Adeegbe, D. & Malek, T. R.
26. Shevach, E. M. Mechanisms of foxp3+ processing and presentation in transplantation. Essential role for interleukin‑2 for CD4+CD25+
T regulatory cell‑mediated suppression. Curr. Opin. Immunol. 8, 681–687 (1996). T regulatory cell development during the
Immunity 30, 636–645 (2009). 43. Rulifson, I. C., Szot, G. L., Palmer, E. & neonatal period. J. Exp. Med. 201, 769–777
27. D’Alessio, F. R. et al. CD4+CD25+Foxp3+ Tregs Bluestone, J. A. Inability to induce tolerance (2005).
resolve experimental lung injury in mice and are through direct antigen presentation. Am. J. 60. Zheng, X. X. et al. Favorably tipping the balance
present in humans with acute lung injury. J. Clin. Transplant. 2, 510–519 (2002). between cytopathic and regulatory T cells to
Invest. 119, 2898–2913 (2009). 44. Hara, M. et al. IL‑10 is required for regulatory create transplantation tolerance. Immunity 19,
28. Joffre, O. et al. Prevention of acute and chronic T cells to mediate tolerance to alloantigens 503–514 (2003).
allograft rejection with CD4+CD25+Foxp3+ in vivo. J. Immunol. 166, 3789–3796 (2001). 61. Webster, K. E. et al. In vivo expansion of T reg
regulatory T lymphocytes. Nat. Med. 14, 88–92 45. Feuerer, M. et al. Lean, but not obese, fat is cells with IL‑2‑mAb complexes: induction of
(2008). enriched for a unique population of regulatory resistance to EAE and long‑term acceptance of
29. Deaglio, S. et al. Adenosine generation catalyzed T cells that affect metabolic parameters. Nat. islet allografts without immunosuppression.
by CD39 and CD73 expressed on regulatory Med. 15, 930–939 (2009). J. Exp. Med. 206, 751–760 (2009).
T cells mediates immune suppression. J. Exp. 46. Graca, L., Cobbold, S. P. & Waldmann, H.
Med. 204, 1257–1265 (2007). Identification of regulatory T cells in tolerated Acknowledgments
30. Wing, K. et al. CTLA‑4 control over Foxp3+ allografts. J. Exp. Med. 195, 1641–1646 X. C. Li and L. A. Turka are supported by the NIH, USA.
regulatory T cell function. Science 322, (2002).
271–275 (2008). 47. Zhang, N. et al. Regulatory T cells sequentially Author contributions
31. Bettni, M. & Vignali, D. A. Regulatory T cells and migrate from inflamed tissues to draining lymph X. C. Li researched data for the article and
inhibitory cytokines in autoimmunity. Curr. Opin. nodes to suppress the alloimmune response. contributed to the discussion of content, writing the
Immunol. 21, 612–618 (2009). Immunity 30, 458–469 (2009). article, and reviewing/editing the manuscript before
32. Collison, L. W. et al. The inhibitory cytokine IL‑35 48. Surh, C. D., Boyman, O., Purton, J. F. & Sprent, J. submission. L. A. Turka contributed to the discussion
contributes to regulatory T cell function. Nature Homeostasis of memory T cells. Immunol. Rev. of content, writing the article, and reviewing/editing
450, 566–569 (2007). 211, 154–163 (2006). the manuscript before submission.

nature reviews | nephrology volume 6 | oCtoBer 2010 | 583

© 2010 Macmillan Publishers Limited. All rights reserved