PerspectiveTransplantation and Autoimmune Disease

Regulatory T Cells: Customizing for the Clinic

See allHide authors and affiliations

Science Translational Medicine  18 May 2011:
Vol. 3, Issue 83, pp. 83ps19
DOI: 10.1126/scitranslmed.3001819


Immune-suppressive cellular immunotherapy requires large numbers of antigen-specific regulatory T cells (Treg cells), lymphocytes that suppress certain immune responses. Together, three papers in this issue of Science Translational Medicine describe protocols for the ex vivo expansion of human Treg cells and assess the immune-suppressive function of ex vivo–manipulated Treg cells after transfer into humanized mouse disease models. Along with recent phase I clinical trial results, these new data provide a platform for clinical use of Treg cells as personalized therapeutic agents for the treatment of autoimmune diseases, graft-versus-host disease, and transplant rejection.

Scientists have displayed both enthusiasm and scepticism—often at the same time—regarding a potential role for immune- suppressing regulatory T cells (Treg cells) in the clinical setting. Treg cells are dedicated suppressors of diverse immune responses and play pivotal roles in the control of immunological self-tolerance as well as immune responses to pathogens and tumor antigens (Fig. 1). Over the past 15 years, researchers have made substantial strides in the understanding of the immunobiology and clinical application of Treg cells in experimental models of transplantation and autoimmune diseases and, recently, in human patients. Harnessing of the immune-suppressive function of Treg cells for cellular immunotherapy (Fig. 2) has been hampered by the low numbers of circulating Treg cells in people. In this issue of Science Translational Medicine, three research groups describe protocols for the generation and ex vivo expansion of antigen-specific human Treg cells and characterize the immune-suppressive function of these cells in humanized mouse models (13). Furthermore, data have become available from recent phase I clinical trials that safety-tested human Treg cells in the context of graft-versus-host disease (GVHD). Together, these new data provide a platform for clinical utility of Treg cells as new individualized therapeutic agents for the treatment of various autoimmune diseases and GVHD and for the induction of solid organ transplantation tolerance.

Fig. 1. Antigen-specific Treg cells: How they suppress.

Activation of antigen-specific Treg cells by their cognate antigen presented by dendritic cells (DCs) induces their suppressive activity. By three different mechanisms, Treg cells suppress effector T cell (cytotoxic T lymphocyte and helper T lymphocyte) functions—such as proliferation, cytokine production, and killing of their antigen-presenting target cells—either directly or via Treg action on effector T cell–stimulating antigen-presenting DCs. (A) First, Treg cells can deliver a negative signal to target cells by direct cell-cell contact, such as (i) interaction of CTLA-4 on Treg cells with DC molecules CD80/CD86, which leads to their down-regulation or inhibition; (ii) interaction of LAG-3 on Treg cells with major histocompatibility complex (MHC) class II molecules on the DCs, which leads to the suppression of DC maturation; (iii) prolonging the interaction of neuropilin-1 (NRP-1) on Treg cells with DCs, which leads to restriction of DC access to effector T cells; (iv) interaction of the serine protease granzyme on Treg cells with effector T cells, which leads to effector cell apoptosis; and (v) interaction of galectin-1 on Treg cells with effector T cells, which leads to effector cell cycle arrest. (B) Treg cells also can deliver a negative signal to target cells by the generation of suppressive soluble factors [IL-10, TGF-β, IL-35 (after TCR activation), or adenosine (derived from extracellular ATP or ADP via its hydrolysis by CD39/CD73] that inhibit effector T cell function or DC maturation. (C) Lastly, Treg cells may also suppress effector T cell function by competing for growth factors (such as IL-2). CD80 and CD86 are T cell costimulatory molecules (ligands for CD28); CD39 is ectonucleoside triphosphate diphosphohydrolase 1; CD73, ecto-5′-nucleotidase; CTLA-4, cytotoxic T-lymphocyte antigen-4; LAG-3, CD4-related lymphocyte activation gene–3; ATP, adenosine 5′-triphosphate; ADP, adenosine 5′-diphosphate.

Fig. 2. Treg cell immunotherapy.

In adoptive transfer of Treg cells for therapy (right), natural Treg cells are isolated from a patient blood sample, expanded to large numbers in vitro, and infused back to the patient. Alternatively, naïve T cells are isolated from the blood of a patient, converted into Treg cells in vitro, and then expanded in vitro to large numbers for therapy. In vivo targeting of Treg cells for therapy (top left) involves the use of monoclonal antibodies (for example, antibodies to CD3, CD40L, CD4, CD28, and CD 52 or CTLA-Ig fusion protein) or pharmacological agents (for example, rapamycin, trichostatin A, IL-2–IL-2 receptor complexes) that activate and expand Treg cells in the patient. On the other hand, in vivo depletion of Treg cells (bottom left) by injection of agents such as anti-CD25 mAb or the diphtheria toxin–IL-2 fusion protein enhances immunotherapy for the treatment of cancer and chronic infections by augmenting antitumor and antiviral effector T cell function.



Once called suppressor T cells (4), Treg cells play a pivotal role in peripheral immunological tolerance, which occurs outside of the thymus and includes hyporesponsiveness in lymphocytes (anergy) and the suppression of self-reactive immune cells. Since 1995, when naturally occurring Treg cells were identified as a minor subpopulation of CD4+ helper T cells that also express CD25 [the α subunit of the receptor for interleukin-2 (IL–2), a T cell growth–promoting cytokine] (5), these cells have been extensively scrutinized (69) and shown to be dedicated suppressor cells that control immunological self-tolerance and immune responses to pathogens and tumor antigens (611). Treg cells are capable of suppressing the effector functions of various cell types, including effector CD4+ and cytotoxic CD8+ T cells, natural killer (NK) cells, B cells, dendritic cells (DCs), macrophages, mast cells, and osteoblasts (611). Once activated, Treg cells also can suppress the response of T cells activated by the same or different antigen-presenting cells (APCs), so-called bystander suppression (611). Compelling data from experimental animal models have indicated that adoptive transfer of Treg cells can prevent or even halt progression of a variety of autoimmune diseases (10, 11) as well as block negative consequences of organ transplantation (GVHD and allograft rejection) (10, 11). On the other hand, in vivo depletion of Treg cells enhances immune responses against infection and cancer (611).

The term Treg includes a diverse group of cells. Some Treg cells develop in the thymus and are known as natural Treg cells (nTreg cells), whereas others are T cells that have been induced, in the periphery, to become Treg cells (iTreg cells) (611). IL-2 signaling is critical for the generation and function of both nTregs and iTregs (611). Both groups consist of various Treg subtypes, which are defined on the basis of phenotypes (10). The most widely studied are CD4+CD25+ Treg cells, which form the key immune-suppressive Treg population (611). In early studies to assess the function of CD4+CD25+ nTreg cells, researchers injected athymic nude mice with a population of CD4+ T cells from nu/+ mice that had been depleted of the CD25+ subset; all recipients spontaneously developed multiple-organ autoimmune diseases, including thyroiditis, adrenalitis, glomerulonephritis, and polyarthritis (5). Reconstitution of CD4+CD25+ T cells 10 days after transfer of the CD4+CD25 T cells suppressed these diseases (5).

A problem for immunologists in the study of Treg cells has been that CD25 is also highly abundant on activated T cells and thus is not a useful discriminating marker for Treg cells. In 2003, the forkhead box transcription factor Foxp3 was identified as a bona fide marker of Treg cells; its expression is necessary and sufficient for regulatory activity, and ectopic transduction of the Foxp3 gene into conventional T cells induces them to become Treg cells (12, 13). Scurfy mice, which lack Foxp3, are deficient in nTreg cells and develop severe lymphoproliferative autoimmune disease (69). Furthermore, mutations in the Foxp3 gene in humans give rise to immune dysregulation, polyendocrinopathy enteropathy X-linked syndrome (IPEX), a life-threatening autoimmune disorder (14).

A complete understanding of the mechanisms of suppression by Foxp3+ Treg cells has yet to be fully elucidated. Foxp3+ Treg cells can suppress effector cells directly or via their action on APCs through three different but not mutually exclusive mechanisms (611, 15) (Fig. 1). The three suppression pathways may operate alone or synergistically, so that each mechanism contributes differentially to suppression in a specific tissue or in a defined inflammatory environment (611, 15).


The use of Treg cells as therapeutic agents (Fig. 2) in adoptive transfer—a mode of treatment in which a patient’s immune cells are modified in vitro and then reintroduced into the person—has several advantages over conventional medications. First, Treg cells have the potential to mediate antigen-specific immunosuppression, thus avoiding the toxicity of general immunosuppressive drugs. Second, Treg cells are generated from the peripheral blood of the patient and thus should not be rejected by his or her immune system. Third, Treg cell therapy can reestablish immunological tolerance and has lasting therapeutic effects in mice. However, to translate Treg cell therapy to the clinic, numerous issues must be addressed regarding human Treg cells. For example, research has indicated that in mice, a Treg may stop being a Treg (16), and Foxp3 expression can be induced in human T cells without necessarily conferring a suppressive function on the cells (17). Therefore, recent studies have focused on deciphering mechanisms that regulate the stability of the Treg cell lineage as well as discovering selective Treg markers in humans and generating large amounts of Treg cells for clinical use.

Human Treg cell heterogeneity. Human Treg cells were initially identified as CD4+CD25+ or CD4+CD25high T cells (18). Subsequently, CD4+CD25+CD127 T cells were reported to be Treg cells because most of these cells express Foxp3 (18). More recently, human Foxp3+ Treg cells were divided into two additional groups based on expression of CD45RA, a transmembrane protein tyrosine phosphatase expressed by naïve T cells: CD45RA+Foxp3low resting Treg cells (rTreg cells) and CD45RAFoxp3high effector Treg cells (eTreg cells). CD45RAFoxp3low cells were classified as non-Treg cells (19).

These three subsets of human Foxp3+ cells display distinct phenotypes and functions. Both rTreg cells and eTreg cells have potent immunosuppressive properties in vitro. rTreg cells can proliferate and differentiate into eTreg cells after T cell receptor (TCR) stimulation, whereas eTreg cells are anergic and prone to die by apoptosis. Interestingly, eTreg cells suppress the proliferation of rTreg cells. CD45RAFoxp3low non-Treg cells produce proinflammatory cytokines IL-17, IL-2, and interferon-γ (IFN-γ) and do not have an immunosuppressive function (18, 19).

In addition to the lineage intricacies of human Treg populations, the observation that human Foxp3+ Treg cells are unstable in vivo complicates their use in adoptive transfer therapies (16, 19). CD45RAFoxp3low non-Treg cells express the retinoic acid receptor–related orphan nuclear receptor γt (RORγt) and, in an inflammatory milieu, can become IL-17–producing T helper cells (TH17 cells), which are believed to play a central role in autoimmune disease (19, 20). Moreover, studies from mice suggest that some Foxp3+ Treg cells may lose Foxp3 expression in vivo as a result of epigenetic modifications of the Foxp3 gene, causing the cells to differentiate into inflammatory memory T cells, which produce inflammatory cytokines such as IL-17 and IFN-γ (16). Therefore, adoptive transfer of unstable therapeutic Treg cells could worsen rather than attenuate disease. It is thus important to use pure populations of Foxp3+ Treg cells and routinely monitor the stability of Foxp3 expression in these cells to ensure the functional quality of Treg cells destined for therapeutic use.

Ex vivo Treg cell expansion. Foxp3+ Treg cells represent less than 10% of CD4+ T cells in human peripheral blood, and large numbers of Treg cells may be required for effective Treg-based cellular immunotherapy (10, 11). Human Treg cells are hypoproliferative in vitro and are prone to undergoing apoptosis after TCR stimulation. It is imperative to design culture conditions in which to expand large numbers of Treg cells ex vivo or to devise protocols for the de novo generation of functional Treg cells in vitro. Small-molecule DNA demethylating agents can maintain stable and high expression of Foxp3 in Treg cells in vitro (69). Foxp3+ Treg cells can also be generated from naïve CD4+ T cells in vitro in the presence of transforming growth factor–β (TGF-β) (611); this process is facilitated by the addition of retinoic acid and the immunosuppressive drug rapamycin by virtue of their ability to inhibit the expansion of IL-17–producing inflammatory TH17 cells and by maintaining stable expression of Foxp3 (21, 22). This approach of de novo Treg cell generation is particularly attractive because some patients with autoimmune diseases have functional defects in their Treg cells, and so these cells cannot be used as therapeutic reagents (10, 11, 18).

Now, two research groups describe crucial technical advances in generating and expanding large numbers of human Treg cells for clinical use. Blazar and colleagues (1) report a protocol with which human peripheral blood CD4+CD25+ Treg cells can be expanded 3000-fold to yield ~600 × 109 cells under good manufacturing practice (GMP) conditions. After isolation from blood, the Treg cells were expanded by culturing in the presence of rapamycin, IL-2, and artificial human APCs derived from the K562 human myelogenous leukemia cell line. The artificial APCs were engineered by means of gene transfer to express the high-affinity antibody-binding Fc receptor (CD64) and the costimulatory molecule CD86 (a CD28 ligand) and were loaded with an antibody to CD3—a part of the TCR; the resulting cells are potent stimulators of T cells, including Treg cells. The expanded Treg cells maintained expression of Foxp3 as well as their in vitro suppressive function (they inhibited the proliferation of CD8+ T cells after TCR stimulation). Importantly, after cryopreservation and restimulation, the expanded human Treg cells significantly reduced GVHD lethality induced by the infusion of allogeneic human peripheral blood mononuclear cells (PBMCs) into immune-deficient mice (1).

Also in this issue of Science Translational Medicine, Wood, Bushell, and colleagues (2) describe the generation of donor-reactive Foxp3+ Treg cells from naïve mouse CD4+ T cells. The cells were stimulated in the presence of allogeneic antigen-presenting DCs (to activate the TCR) and the phosphodiesterase 3 (PDEi) inhibitor cilostamide, which increases intracellular adenosine 3′,5′-monophosphate (cAMP) by blocking its hydrolysis; this enrichment of cAMP is thought to drive the development of functional Treg cells. The resulting cell population blocked skin allograft rejection in mice. The authors then carried out the same procedure with human PBMCs, and the resulting CD4+Foxp3+ Treg cells attenuated chronic transplant vasculopathy after adoptive transfer in a humanized mouse model of artery transplantation (2). Foxp3+ Treg cells express, on the external faces of their cell membranes, CD39 and CD73, enzymes that hydrolyze adenosine 5′-triphosphate (ATP) and adenosine 5′-diphosphate (ADP) to yield extracellular free adenosine for Treg suppressive functions (Fig. 1).

Another active area of research is the use of antigen-specific Foxp3+ Treg cells in adoptive cellular therapy. Mouse studies have shown that antigen-specific Foxp3+ Treg cells are more effective than polyclonal Treg cells in suppressing autoimmune diseases, including diabetes, in nonobese diabetic (NOD) mice (10, 11). Several approaches have been reported to generate and expand antigen-specific human Treg cells (10, 11). One such approach is to expand the cells ex vivo in the presence of DCs loaded with a peptide derived from the patient’s histocompatibility leukocyte antigen A2 (HLA-A2 antigen). The antigen specificity of the resulting Treg cells can be verified by binding of their TCRs to the specific peptide conjugated to an HLA-DR tetramer (23), a reagent used to detect rare antigen-specific T cells. Another method is to use gene transfer of antigen-specific TCR Vα and Vβ chains that recognize a specific peptide antigen into polyclonal Foxp3+ Treg cells. These antigen-specific Treg cells express tissue-specific homing receptors, leading to site-specific regulation of inflammatory arthritis and organ transplant rejection in mice (24, 25).

Now, Sagoo et al. report a new method by which to enrich for donor alloantigen–specific human Treg cells after stimulation with allogeneic DCs. The enrichment method was based on the co-expression of two Treg activation markers, CD69 and CD71 (3), and the cell population was then expanded in vitro. The expanded alloantigen-specific Treg cells recognized intact HLA-DR antigens, and the suppressive functions of these cells were tested in a humanized mouse model of immune-related injury to human skin grafts; the expanded alloantigen- specific human Treg cells were shown, by a number of clinical measures (apoptosis, epidermal keratinocyte proliferation, and microvessel injury of CD31+ endothelial vessels in the allografted tissue), to better protect the allogenic grafted tissue from injury, relative to polyclonal Treg cells. The allografts of the antigen-specific Treg–treated group also contained a higher proportion of CD3+Foxp3+ cells relative to the control group. These findings augment data from previous studies that support the hypothesis that antigen-specific Treg cells perform better as therapeutic reagents than do polyclonal Treg cells (10, 11, 2325).

Because of the heterogeneity of human Foxp3+ Treg cells, to date there is no consensus as to which Treg population is optimally suitable for clinical use. The data presented in this issue of Science Translational Medicine are encouraging, but it remains to be seen how well these customized approaches to the generation of human Treg cells translate to clinical use.

Animal models of disease have shown that adoptive transfer of Treg cells is effective only during the autoimmune disease– induction phase and not in the advanced stages of disease (10, 11). These observations suggest that the efficacy of Treg cell therapy could be enhanced by depleting tissue-damaging effector T cells or by reducing the numbers of these cells to a level that is manageable by Treg cells. This combined therapeutic strategy may have additional advantages because some effector T cell–depleting agents are also capable of activating and expanding Treg cells in vivo (Fig. 2).


A number of pharmacological agents have been reported to be capable of activating and expanding Treg cells in mouse models of disease. These include the immunosuppressive cytokines TGF-β (induces Foxp3+ Treg cells and TH3 cells) and IL-10 [induces Foxp3+ Treg cells and a distinct suppressor population, type 1 regulatory T (TR1) cells, which promote tolerance to foreign and some self-antigens] (10, 11) and the histone deacetylase inhibitor trichostatin-A, which induces Foxp3 expression (26). Agents such as the immunosuppressive drug rapamycin and antibodies to CD3, CD4, CD8, or CD40L (CTLA-Ig fusion protein) both inhibit T cell activation and selectively activate and expand Treg cells in vivo; some of these are being tested in the clinic, including rapamycin for allograft rejection, the CTLA-Ig fusion protein for rheumatoid arthritis and allograft rejection, and a humanized antibody to CD3 for type 1 diabetes (2729) (Fig. 2).

It has been reported that tumor cells escape host immune responses by recruiting Treg cells (30). Thus, depletion of Treg cells or abrogation of their suppressive function may enhance host immunity against cancer. One strategy that has been successful in mouse models is to use anti-CD25 monoclonal antibodies (mAbs) to eliminate Treg cells in vivo. However, activated effector T cells also express CD25, and anti-CD25 mAbs also deplete tumor-specific effector T cells. On the other hand, the diphtheria toxin–IL-2 fusion protein seems promising (Fig. 2) because it binds to CD25 on Treg cells and, after internalization, inhibits protein synthesis and induces Treg cell death in patients (31). A recent phase II trial of the diphtheria toxin–IL-2 fusion protein in 16 patients with stage IV melanoma revealed that this reagent is effective in inducing regression of tumor metastases. The reagent was used to transiently deplete CD25+ Treg cells, allowing the repopulation of tumor-specific cytotoxic CD8+ effector T cells, which blocked tumor progression (31).


To date, several phase I clinical trials have tested the ability of ex vivo–expanded CD4+CD25+ Treg cells to prevent GVHD after allogeneic bone marrow transplantation (Table 1), and many other trials are under way. One study, led by Blazar and colleagues, used cryopreserved CD4+CD25+ Treg cells —isolated from partially HLA-matched third-party umbilical cord blood (UCB) and expanded in vitro—to treat GVHD in 23 cancer patients who had received UCB transplantations (32). The conditioning regimen consisted of chemotherapy with the DNA alkylating agent cyclophosphamide and DNA synthesis inhibitor fludarabine; total body irradiation (to suppress the patient’s immune system); adoptive transfer of CD4+CD25+ Treg cells (0.1 × 105 to 30 × 105/kg); and immunosuppression [with cyclosporine/mycophenolate mofetil (MMF) in 17 patients and rapamycin/MMF in six patients]. In these patients, Treg cells could be detected for 14 days after infusion, and the incidence of grade II to IV acute GVHD was reduced in the Treg-treated patients relative to historical controls (43 versus 61% in historical controls, P = 0.05), with no deleterious effects on risk of infection, relapse, or early mortality (32).

Table 1. Current phase I clinical trials for Treg cell therapy. Ref, reference.
View this table:

A clinical trial led by Martelli and colleagues used freshly isolated CD4+CD25+ Treg cells from donor blood to treat GVHD in 28 cancer patients (33). The conditioning regimen consisted of total body irradiation, chemotherapy with fludarabine and the DNA alkylating agents thiotepa and cyclophosphamide, and adoptive transfer of CD4+CD25+ Treg cells (2 × 106/kg). In this trial, no posttransplantation immunosuppression was used. Twenty-six of the patients achieved primary, sustained full–donor type engraftment, and two developed grade II acute GVHD; these data could not be compared with historical controls because all of those patients received immunosuppression treatment after bone marrow transplantation. At a median follow-up of 11.2 months (range 3.6–21.4), no patient has developed chronic GVHD (33).

In addition to these clinical trials, other ongoing studies indicate that adoptive Treg cellular therapy is proving to be safe and effective to treat many forms of disease (Table 1) (3235). In the planning stages are new clinical trials to test the safety and effectiveness of Treg cell therapy in individuals with autoimmune diseases such as type 1 diabetes and in solid organ–transplantation patients. One of the most serious potential side effects of Treg cell therapy is the inhibition of antitumor and antiviral immunity. These concerns were not addressed in the phase I trials described herein and can only be assessed in larger-scale human trials. As the techniques required to culture large numbers of antigen-specific human Treg cells emerge, we expect that in the next few years patient-specific Treg cell therapy will be advanced into the clinic as individualized treatments for autoimmune diseases and GVHD and for the induction of solid organ–transplantation tolerance.


  • Citation: X. Wang, L. Lu, S. Jiang, Regulatory T Cells: Customizing for the Clinic. Sci. Transl. Med. 3, 83ps19 (2011).

References and Notes

  1. Competing interests: The authors declare no competing interests.
View Abstract

Stay Connected to Science Translational Medicine

Navigate This Article