Promoting transplantation tolerance; adoptive regulatory T cell therapy

Abstract

Transplantation is a successful treatment for end-stage organ failure. Despite improvements in short-term outcome, long-term survival remains suboptimal because of the morbidity and mortality associated with long-term use of immunosuppression. There is, therefore, a pressing need to devise protocols that induce tolerance in order to minimize or completely withdraw immunosuppression in transplant recipients. In this review we will discuss how regulatory T cells (T_regs) came to be recognized as an attractive way to promote transplantation tolerance. We will summarize the preclinical data, supporting the importance of these cells in the induction and maintenance of immune tolerance and that provide the rationale for the isolation and expansion of these cells for cellular therapy. We will also describe the data from the first clinical trials, using T_regs to inhibit graft-versus-host disease (GVHD) after haematopoietic stem cell transplantation and will address both the challenges and opportunities in human T_reg cell therapy.

OTHER ARTICLES PUBLISHED IN THIS SERIES

T cell depletion in paediatric stem cell transplantation. Clinical and Experimental Immunology 2013, 172: 139–47.

Tolerogenic dendritic cell therapy for rheumatoid arthritis: where are we now? Clinical and Experimental Immunology 2013, 172: 148–57.

Whole body imaging of adoptively transferred T cells using magnetic resonance imaging, single photon emission computed tomography and positron emission tomography techniques, with a focus on regulatory T cells. Clinical and Experimental Immunology 2013, 172: 169–77.

Introduction

Advances in surgical techniques and the introduction of T cell-directed immunosuppressive agents has made solid organ transplantation a well-established treatment for end-stage failure of several major organs. Despite improvements in short-term outcome, long-term patient and graft survival remain suboptimal due to the toxic side effects associated with long-term use of these drugs. A major goal of transplantation research is, therefore, to promote ‘tolerance’, a state in which the host's immune system can be reprogrammed and then guided to accept a transplant without the need for long-term immunosuppression. In this pursuit, clinically applicable protocols aim to tip the balance in favour of regulation by either the in-vivo expansion of T cells with regulatory activity or the infusion of ex-vivo expanded cells.

The past two decades have seen the discovery of many different types of regulatory T cells, including: CD8⁺ T cells 1, CD4^–CD8^– double-negative T cells 2, CD8⁺CD28^– 3, natural killer (NK) T cells 4 and γδ T cells 5, but these are less well studied compared to CD4⁺ regulatory T cells (T_regs). In this review we will focus on the potential for clinical application of CD4⁺ T_regs, characterized by high and stable expression of surface interleukin (IL)-2 receptor α chain (IL-2Rα, CD25^hi) and the transcription factor, forkhead box protein 3 (FoxP3) 6. These CD4⁺CD25⁺FoxP3⁺ cells are thymus-derived, referred to as natural T_regs (nT_regs), compared to their counterparts that are generated in the periphery and whose activation requires T cell receptor engagement and cytokines, the induced T_regs (iT_regs) 7,8.

In comparison to iT_regs, studies support the more potent and stable role of nT_regs (referred to hereafter as T_regs) in maintaining self-tolerance and preventing autoimmunity 9. The ability to expand such cells has, therefore, become an attractive prospect in modulating immune responses not only in the context of solid organ transplantation, but also in autoimmunity and prevention of graft-versus-host disease (GVHD). The rationale is based on animal models and clinical studies that have demonstrated clearly that T_reg deficiency and/or functional defects might contribute to the pathophysiology of several autoimmune diseases such as type I diabetes, multiple sclerosis, rheumatoid arthritis, as well as organ rejection (reviewed in 10).

In the context of organ transplantation, it is of paramount importance to understand the way in which alloreactive CD4⁺ T cells see alloantigen in order to better dictate the strategies used for the clinical application of T_regs. The two main pathways of allorecognition include the direct pathway [graft-derived cells are recognized by their surface expression of intact major histocompatibility (MHC) alloantigens], and the indirect pathway [recipient antigen-presenting cells (APC) present donor MHC alloantigens in the context of self-MHC 11]. Direct allorecognition is a vigorous reaction due to the high precursor frequency of alloreactive T cells; in this regard it is generally accepted that deletion of a substantial proportion of direct pathway alloreactive T cells will be required to ‘tip the balance’ from reactivity to regulation 12,13. In addition, in order to suppress the surviving alloreactive T cells by regulation one would need sufficient numbers of T_regs in the right place, at the right time, in an environment that favours regulation. Therefore, the specificity of the T_regs chosen for cellular therapy may play an important role (discussed in later sections). The main focus of this review is the clinical application of T_regs in the setting of transplantation and the journey from bench to bedside. We will discuss the challenges that we still face in the laboratory from the isolation to the ex-vivo expansion of these cells for immunotherapy and outline the questions that still remain with regard to the clinical protocols. Moreover, human T_regs are currently less well-characterized and understood compared to mouse T_regs; we will, therefore, review briefly their biology before discussion of their clinical application.

Regulatory T cells (T_regs)

Markers used for isolation

Aside from the expression of CD25 14 and FoxP3 (outlined above), human T_regs also express CD27 15, CD45RA 16, CD39 17, CD122, cytotoxic T lymphocyte antigen-4 (CTLA-4 or CD152) and the glucocorticoid-induced tumour necrosis factor receptor (GITR) family-related gene 18,19. However, most of these cell surface markers are not exclusive to T_regs, with some of these markers also expressed by non-regulatory CD4⁺ T cells, posing a challenge during the isolation process. As an example, data support the key role of FoxP3 in the development, maintenance and function of T_regs with supporting evidence that point mutations in the FoxP3 gene leads to a functional T_reg deficit that is evident in patients with IPEX (immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome) 20. Despite this, FoxP3 is not a sufficient marker for the isolation of T_regs, as many activated effector T cells also express FoxP3 without having a regulatory phenotype 21. Moreover, being an intracellular protein, this marker cannot be used to isolate T_regs. What complicates the story even further is that human T_regs are heterogeneous. In contrast with mice, the combination of the marker CD45RA and the level of expression of FoxP3 delineates the human T_reg compartment into naive or resting T_regs (CD45RA⁺FoxP3^low), effector T_regs (CD45RA^–FoxP3^high), both of which are suppressive in vitro, and the non-suppressive, cytokine secreting non-T_regs (CD45RA^–FoxP3^low) 22,23.

The search for further surface markers to aid in the unique identification of T_regs has led to studies showing that T_regs also express low levels of CD127, the α-chain of the IL-7 receptor 24. Accordingly, the constitutive high expression of CD25 but low expression of CD127 has been used to discriminate T_regs from activated effector T cells 25. However, the combination of CD25 and CD127 is still not sufficient to isolate functionally pure T_regs, bearing in mind that not all the ex vivo-isolated FoxP3⁺ T_regs are regulatory.

Such studies, therefore, highlight the fact that despite all the efforts to identify T_reg markers, the quest continues and we have yet to find markers that define ‘pure’ T_reg populations for the purposes of cellular therapy.

Mechanism of action

Several mechanisms of suppression by T_regs have been proposed. T_regs can suppress the functional ability of both CD4⁺ and CD8⁺ T cells directly by preventing their differentiation, activation and proliferation via either cell–cell contact or a contact-independent route, which includes inhibitory cytokines such as IL-10, transforming growth factor (TGF)-β and recently IL-35 26–28. They can also kill effector T cells directly in a perforin-dependent and granzyme-dependent manner or suppress their activation 29,30. Furthermore, T_regs have been shown to express galectin-1, with blockade of galectin-1 binding to activated T cells being shown to reduce the T_reg inhibitory effect 31.

Moreover, T_regs may mediate their suppressive function by acting directly on dendritic cells (DCs), attenuating their antigen-presenting and co-stimulatory functions. In support of this, Fassbender et al. 32 showed that the co-culture of murine DCs with T_regs led to an increase in DC cyclic adenosine monophosphate (cAMP), which was responsible for the down-regulation of the co-stimulatory molecules, CD80/CD86. Other mechanisms include the role of cytotoxic T lymphocyte antigen 4 (CTLA-4), a negative co-stimulatory molecule on T_regs, in either up-regulating indoleamine 2, 3-dioxygenase (IDO) expression on DCs which, in turn, down-regulated immune responses 33, or acting as an effector molecule to inhibit CD28 co-stimulation by the cell-extrinsic depletion of co-stimulatory ligands 34.

As evident from the studies outlined, therefore, it becomes clear that the precise mechanism of suppression by T_regs has yet to be fully elucidated.

Preclinical data in support of adoptive T_reg cell therapy

The term ‘adoptive immunity’ was first coined in 1954 by Billingham et al. 35, who were able to show that passive transfer of primed immune cells can generate immunity in the recipient. Subsequently, numerous animal studies have demonstrated the effectiveness of this adoptive transfer of immunity towards cancer and infectious disease 36,37. Moreover, the use of IL-2 permitted, for the first time, the ex-vivo culture and expansion of T cells in humans 38. In addition, many transplant researchers found that CD4⁺ T cells were responsible for donor-specific tolerance, and it was the study by Hall et al. 39 which concluded that transplant tolerance was mediated by CD4⁺CD25⁺ cells. In this study they showed that in cyclosporin-treated rats with long-term cardiac allograft survival, the adoptive transfer of CD4⁺CD25⁺ T cells resulted in tolerance.

The application of T_regs in the context of organ transplantation is supported further by the seminal work by Sakaguchi et al. 6, who showed that T_regs from naive mice prevented rejection of allogeneic skin grafts in T cell-deficient nude mice given CD25^– T cells.

Subsequently, a series of preclinical rodent models of skin and cardiac transplantation demonstrated that T_regs present in the recipient at the time of transplantation are critical in the induction and maintenance of tolerance (reviewed in 40). In support of such studies we have also generated T_reg lines in vitro, and shown that these T_regs are very effective at inducing survival of MHC-mismatched heart allografts 41. Furthermore, in a murine skin transplant model following thymectomy and partial T cell depletion, we have demonstrated previously the ability of in-vitro-expanded T_regs in inducing donor-specific transplantation tolerance in this system 42.

The importance of adoptive T_reg therapy in transplantation is supported further in mouse models of bone marrow transplantation, where the transfer of freshly isolated T_regs together with the bone-marrow allograft has been shown to ameliorate GVHD and facilitate engraftment 43. GVHD was also the first model in which it was shown that the adoptive transfer of ex-vivo-expanded donor T_regs was highly effective in preventing acute or chronic GVHD 44. Moreover, the adoptive transfer of T_regs has been shown to prevent rejection of pancreatic islet 45 and other organ allografts 46,47.

The use of currently available humanized mouse models of GVHD and allotransplantation 48,49 has reinforced further the importance of T_regs in these settings. These models are based on the reconstitution of immunodeficient mice with human immune cells. More recently we have also shown the efficacy of human T_regs in preventing alloimmune dermal tissue injury in a humanized mouse model of skin transplantation 50. Furthermore, Nadig et al. 51 developed a human vessel graft model to study the in-vivo function of T_regs. Their results showed convincingly that grafts from mice reconstituted with peripheral mononuclear cells (PBMCs) alone exhibited extensive vasculopathy, whereas the co-transfer of T_regs prevented this process.

Such adoptive transfer experiments in rodents, therefore, support the notion that tolerance requires ‘tipping the balance’ between reactivity and regulation. Despite such data generated in preclinical animal models, showing successfully that T_regs can induce and maintain transplantation tolerance, we currently face many challenges in the laboratory that have hindered the widespread application of T_reg cell therapy in the transplant setting. In addition, a number of different strategies have been proposed for the isolation and expansion of T_regs for cellular therapy. However, there is no consensus on the optimal process, and many such processes have their limitations (discussed below).

Clinical application of human T_regs: challenges with large-scale manufacture

Isolation and expansion

One of the obstacles in the implementation of clinical protocols using T_regs is their low frequency, 1–3% of total peripheral blood CD4⁺ T cells, and data (from animal models) which suggest that, for these cells to suppress immune responses, high doses of T_regs in relation to effectors is required 52,53. This means that for cellular therapy, it will almost certainly be necessary to use a polyclonal stimulus to expand T_regs in vitro. In this regard, the large-scale ex-vivo expansion of human T_regs by stimulation with anti-CD3 and anti-CD28 monoclonal antibody-coated beads and high-dose IL-2 has been demonstrated successfully 54. However, effectors have the potential to proliferate vigorously under such conditions, so that even a trace of effectors in the starting population can be expanded in high numbers. The injection of such cells would, therefore, be detrimental to the patient and may lead to rejection. Thus, it is essential to either initiate the expansion culture with highly purified T_regs (a challenge in view of the absence of a T_reg-specific cell surface marker) or create culture conditions that favour T_reg cell growth.

Two different combinations of markers appear to be promising for T_reg isolation. The first seeks to isolate CD4⁺CD25^hi T_regs, but with the addition of an antibody to select for CD45RA⁺ cells and so eliminate antigen-experienced or memory T cells 16. The second combination also uses the CD4⁺CD25^hi phenotype, but includes CD127 expression. The rationale for using CD127 as a marker for T_reg isolation (as explained in earlier sections) is on the basis that in human T_regs there is a reciprocal expression of CD127 and FoxP3, and thus CD127 provides a sortable surrogate marker for FoxP3⁺ T_regs 24. Moreover, the so-called ‘naive’ T_reg population based on the co-expression of CD4 and CD45RA yield T_regs with a greater suppressive capacity than total CD25^hi cells 55. The reason for this became clear when Miyara et al. 22 noted the subpopulations of human FoxP3⁺ T cells and discovered that the CD25⁺CD45RA^-FoxP3^hi cells contain many Th17 precursors. Furthermore, after 3 weeks of in-vitro expansion the CD45RA⁺-expanded T_regs remained demethylated (compared to the CD127^– T_regs that became methylated) at the T_reg-specific demethylation region (TSDR), which is a conserved region upstream of exon 1 within the FoxP3 locus [completely demethylated in natural T_regs but methylated fully in induced T_regs and effector T cells (T_eff)] 55,56. Such studies, therefore, support the isolation of T_regs based on CD45RA⁺ expression, bearing in mind that they are the most stable population for expansion and have the greatest expansion potential 16. Despite such studies, one drawback is that the number of naive T_regs declines in the peripheral blood with age 57, and hence isolation based on CD127 expression may still be a practical approach. Moreover, two elegant studies by Kathryn Wood's group 51,58 support the in-vivo superiority of the CD4⁺CD25⁺CD127^low T_regs at regulating alloreactivity compared to T_regs isolated based on the expression of CD4 and CD25 alone. Such studies have important implications for the design of future clinical studies.

The search for further surface markers to aid the isolation of purer or more potent T_reg populations led to studies investigating markers such as CD121a/CD121b, TGF-β/ latency associated peptide (LAP) 59 and CD39 60. However, all these proteins are expressed only on activated T_regs and would be of use only to re-isolate T_regs after expansion. This may not be feasible, in view of the costs of re-isolating billions of T_regs on a per-patient basis. Other studies complicate the story even further. Ito et al. 61 showed that FoxP3⁺ T_regs could be grouped into two subsets based on the expression of the inducible T cell co-stimulator (ICOS). They showed that while ICOS^–FoxP3⁺ T_regs mediate their suppressive function via TGF-β, ICOS⁺FoxP3⁺ T_regs additionally secrete IL-10. Therefore, depending on the type of immune response to be suppressed, it may be useful to isolate subsets of T_regs which have specific mechanisms of action. Moreover, a recent study by Ukena et al. 62 compared different T_reg isolation strategies in order to define the most promising T_reg target cell population for cellular therapy. They compared CD4⁺CD25^hi enrichment, CD4⁺CD25^hi enrichment and depletion of CD127⁺, enrichment of CD4⁺CD25^hiCD45RA T cells, depletion of CD49d⁺ (a marker of proinflammatory cytokine-producing effector T cells) and CD127⁺ T cells and enrichment of CD4⁺CD25^hi ICOS⁺ and ICOS^– T_regs. They concluded that while CD4⁺CD25^hiCD127^– and CD4⁺CD25^hiICOS⁺ T_regs are the most promising T_regs for fresh cell infusions in clinical trials with respect to cell yield, phenotype, function and stability, the CD4⁺CD25⁺ T_regs qualify as the best candidate for in-vitro expansion. Such studies, therefore, paint a complicated picture that when choosing the T_reg marker for cell isolation we should also bear in mind other factors other than simply purity, i.e. isolating potent cells with a mechanism of action to suppress the immune response of interest and cells with the desired expansion profiles.

Despite this, however, what limits choice when devising a clinically applicable protocol is that isolation techniques need to be good manufacturing practice (GMP)-compliant, and GMP purification reagents for all the various markers outlined above are not yet available.

The clinical T_reg selection protocols used to date in the United Kingdom have used a combination of depletion and positive selection steps, with the isolation tools involving mainly the automated CliniMACS plus system (Miltenyi Biotec, Bisley, UK). This enables GMP-compliant cell selection by magnetic bead activated cell sorting 63. More specifically, for the production of CD4⁺CD25^hi T_regs the GMP-grade antibodies available enable the depletion of CD19/CD8-expressing immune cells followed by a positive selection of CD25-expressing T cells 64. The major drawback with such techniques is that this process does not guarantee the selection of CD25^hi cells compared to the fluorescence activated cell sorter (FACS) sorter, which allows the important distinction to be made between the CD4⁺CD25^hi and CD25^int cells. In addition, the process does not allow the selection of T_regs based on multiple parameters and the ∼60% purity of the isolated cells 65 is not comparable with the >95% purity achieved using the FACS sorter 56.

In addition to the automated CliniMACS plus system (Miltenyi Biotec), there are two other commercially available methods for GMP-grade T cell isolation and expansion. Life Technologies Ltd (Paisley, UK) produces the DynaMag™ CTS™ system, which is a magnetic device used in combination with the Dynabeads® CTS™ and Dynabeads® ClinExVivo™ to positively isolate bead bound cells or deplete unwanted cell types. Dynabeads® CD3/CD28 CTS™ are used to positively isolate T cells; these beads are also able to activate the bound T cells and when cultured in the presence of IL-2 result in a 100–1000-fold expansion of the isolated T cells. The T cells are purified by labelling cells with mouse immunoglobulin (Ig)G1 antibodies and using the Dynabeads® IgG1 Binder CTS™ for positive isolation, negative isolation or cell depletion.

Stage Cell Therapeutics (Göttingen, Germany) is a cell therapy company that manufactures Streptamer® reagents for isolation of defined lymphocytes. In view of isolating purer T_reg populations, their system involves three positive selection steps by magnetically tagged Fab-Streptamers. Following each labelling and positive selection step, the tagged cells are liberated completely from the magnetically tagged Fab-Streptamers by incubation with a competing Streptactin ligand D-biotin that causes disruption of the Fab-multimer complex, dissociation of the Fab-Streptamer label from the target cell surface and complete removal upon washing. The first positive isolation step involves anti-CD4-Fab-Streptamer labelling, followed by anti-CD25-Fab-Streptamer labelling, and finally anti-CD45RA-Fab-Streptamer labelling is used to isolate a triple-positive T_reg cell preparation that is CD4⁺CD25⁺CD45RA⁺.

Interestingly, however, the study by Marek et al. 66 showed that regardless of the initial phenotypic markers used for isolation (i.e. CD25^hiCD127^low, CD45RA⁺, CD45 RA^–) during the expansion process, T_regs were transforming into effector/memory-like cells which produced inflammatory cytokines. They proposed that independent of the phenotypic markers used for T_reg isolation, the only variable to help maintain the T_reg phenotype and function was limiting the expansion time to 2 weeks. Based on such studies, therefore, it is of particular importance to ensure that the stability of the T_regs is maintained during the expansion process.

Basu et al. 67 described a method for the expansion of T_regs in the presence of the immunosuppressant, rapamycin (sirolimus). They showed that rapamycin inhibits preferentially the proliferation and function of CD25⁺ conventional effector T cells and thus permits the expansion of T_regs even from a mixed starting population 67,68.

Furthermore, and in support of such a study, Tresoldi et al. 69 showed that only the expansion cultures in the absence of rapamycin are contaminated by the CD4⁺CCR6⁺CD161⁺ T helper type 17 (Th17) precursor cells. Despite this promise, adding rapamycin to T_reg cultures has its own disadvantages in view of diminishing overall T_reg expansion 70. The addition of rapamycin may, therefore, necessitate extended expansion times in order to achieve the therapeutic numbers – a problem, bearing in mind studies showing loss of FoxP3 expression in human T_regs upon repetitive stimulation (mentioned earlier 55). It is also important to consider that target doses of expanded T_regs may not always be reached, as reported in a clinical trial by Brunstein et al. 71, even when using protocols without the addition of rapamycin. Such trials used anti-CD3/CD28 beads for stimulation and expansion of the T_reg lines, the only GMP reagents available (with a safety record in humans). However, stimulation with cell-based artificial APCs (aAPCs), expressing the co-stimulatory molecule CD86 and an Fc receptor (FcR) for loading of anti-CD3 monoclonal antibody (mAb), has also been used to expand T_regs 72 with approximately fourfold superiority over the use of anti-CD3/CD28 beads.

These studies, therefore, highlight the many obstacles that we still need to overcome to refine further the current protocols for the isolation and expansion of T_regs to ensure safe and efficacious application in the clinical setting. Despite these hurdles in the laboratory, there is still much debate over the specifics of the clinical protocol (outlined below).

Clinical protocol

Immunosuppression and timing of T_regs

Most transplant recipients are treated with a combination of immunosuppressive drugs and biological agents to control rejection and/or GVHD responses. The combination of drugs used varies depending on the type of organ being transplanted as well as the protocols used by individual transplant centres. For example, some countries use induction therapy with monoclonal or polyclonal antibody preparation such as alemtuzumab or anti-thymocyte globulin (ATG) at the time of transplantation. This treatment markedly depletes most of the leucocyte populations in the peripheral blood. Interestingly, leucocyte depletion has the potential to tip the balance in favour of immune regulation by creating a situation whereby regulatory immune cells outnumber the effector cells. However, whether or not induction therapy is used, when devising clinical protocols to incorporate T_regs it is crucial to take into account the influence of the various immunosuppressants on the T_regs in vivo.

Taking the example of ATG, the data indicate that in both adults and children ATG causes a rapid decrease in naive CD4 and CD8 T cells which persist usually for at least 2 months 73. In addition, it has been shown that treatment with ATG is associated with the expansion of FoxP3⁺ T cells in vivo and suggests a shift in T_reg to a T_eff ratio. Despite this, CD4⁺ and CD8⁺ memory cells are resistant to depletion by ATG and these cell subsets expand over the initial 6 months post-transplantation 73. The fact that memory cells survive deletion may explain why patients do not suffer opportunistic infections post-ATG therapy. However, these cells can contribute to early graft injury and loss and, importantly, these cells are more resistant to suppression by T_regs than naive T cells 74. However, to limit memory T cell expansion (post-induction therapy), transplant recipients are maintained on other immunosuppressive drugs, most commonly a calcineurin inhibitor (CNI) such as tacrolimus or cyclosporin A, and an anti-proliferative agent such as mycophenolate mofetil. It has been proposed that both types of drug inhibit the generation and function of T_regs. Despite this, in animal models in the context of autoimmunity it has been shown that for T_regs to exert their suppressive function tissue inflammation needs to be controlled 75. It seems that for T_regs to expand in vivo and exert their suppressive function they require a tolerogenic milieu. In support of this, a recent study analysing the dynamics of the alloimmune response in vivo demonstrated a rapid invasion of effector cells in the grafts followed by the delayed arrival of T_regs that were ineffective at controlling tissue damage 76. In contrast, when the recipient mice were treated with anti-CD40L mAb and rapamycin, effector T cell infiltration was delayed and more than 30% of the graft infiltrating T cells were T_regs. Of note, there is good evidence in the literature indicating that rapamycin is superior to tacrolimus for the thymic export and survival of T_regs 77,78. In contrast to CNIs, rapamycin appears to be tolerance-permissive by selectively inducing apoptosis or necrosis of alloreactive effector cells while promoting T_reg induction 79, expansion 78 and function 80. This may suggest that rapamycin is the ideal candidate for short-term therapy post-depletion in humans. However, rapamycin monotherapy post-depletion is associated with a high risk of acute rejection 81, and it is not yet clear whether the concomitant therapy with T_regs would be sufficient to prevent this or whether further immunosuppression will be required in the short term.

The use of combinations of immunosuppressive agents in the clinical setting highlight the challenge associated with designing protocols that include the infusion of T_regs. Thus, the competing actions of each immunosuppressive drug may have to be considered together with the key question of the timing of cell injection.

Most of the preclinical T_reg therapy studies published thus far administered T_regs prior to or at the time of transplantation, in contrast to studies of autoimmune disease where it has been shown that T_reg therapy after disease initiation can effectively reverse the disease course 82. The intensity of the anti-allograft response and the fragility of the transplanted organ may explain the lack of efficacy when T_reg infusion is delayed. However, if T cell-depleting reagents such as ATG are used as induction therapy, it may be possible to delay T_reg infusion until lymphocyte numbers start to recover 2 months or more after transplantation. This might tip the balance between T_regs and T_eff cells and help to promote a tolerant state.

Location of T_reg function and site of T_reg injection

An additional consideration regarding T_reg therapy is the site of action of T_regs and, consequently, the desired homing properties of injected cells. In the transplant setting, T_reg lymph node homing and their ability to traffic to grafts are both required for their protection against graft rejection 83. Interestingly, in a mouse islet transplant model, therapeutic T_regs function initially at the graft site (preventing the exit of donor-derived DCs) and then traffic to the draining lymph node and continue to exert their suppressive function there 84. In so doing, they prevent the exit and migration of donor-derived DCs to the lymph nodes, thereby reducing alloimmune priming. The translation of such a study to the clinic may mean that to ensure that T_regs exert their suppressive function we need to either inject the cells at the graft site or ensure that the cells reach the graft/lymph node due either to their alloantigen specificity or homing receptor expression. Bearing in mind the serious complications associated with injection of the cells at the graft site, i.e. the risk of bleeding if cells are injected via the portal vein (in the case of liver transplantation), the favoured option is infusion via a peripheral vein.

Studies have shown antigen-specific T_regs to be more potent than polyclonally activated T_reg cells 85–87. Moreover, T_regs with direct specificity are very potent in preventing acute rejection early after transplantation, while T_regs with indirect specificity seem to be crucial to prevent chronic rejection 42,46. In addition, using antigen-specific T_regs would have additional advantages; first, their action would be limited to the site of alloantigen source and immune activation 88,89; and secondly, this may avoid the undesirable pan-suppression, mediated by polyclonal cells, resulting in an increased risk of infections and cancers. However, although the expansion of direct pathway allospecific human T_regs has been achieved 90,91, expansion of indirect pathway T_regs has proved more difficult, posing further challenges 92,93. It is important to note, however, that despite the evidence supporting the importance of antigen-specific T_regs in preventing solid organ rejection, in GVHD the transfer of T_regs enriched for alloantigen specificity showed only moderately improved efficacy when compared to the polyclonal T_reg cell population 94. Although such studies emphasize the lack of antigen-specific requirement for the transferred T_regs, interestingly, a recent study discussed the importance of homing receptor expression in this transplant setting.

Ukena et al. 95 showed that tolerant patients without GVHD after haematopoietic stem cell (HSC) transplantation expressed significantly higher levels of the chemokine receptors transplantation. This may suggest that homing of T_regs to secondary lymphoid tissue and sites of inflammation may play an important role in the control of GVHD, despite some studies suggesting that GVHD is a systemic disease and the concentration of T_regs at a localized site is not required.

These types of study, therefore, support the notion that therapeutic strategies using T_regs have to take into account the fact that these cells not only need potent suppressive function, but also need appropriate tissue trafficking to enable contact with their target cells. Therefore, if the T_regs are to be injected via a peripheral vein then it is important that they express the molecules such as CD62L and CCR7 that are crucial for their migration to the lymph nodes and other chemokine receptors, e.g. CXCR3 for liver homing 96. Moreover, T_regs vary in their expression of trafficking and homing receptors according to their individual histories and state of activation. They have been shown to variously express CCR2, CCR4, CCR7, CCR8, CCR9, CXCR1 and CXCR4 (reviewed in 97). In addition, it is now known that within the pool of FoxP3-expressing cells functionally diverse T_reg subsets can be identified on the basis of chemokine receptor expression 98. In view of the importance of T_reg expression of chemokine receptor and trafficking on their in-vivo suppression function, efforts have been made at understanding the influence of culture conditions on the expression pattern of these receptors on T_regs. In this regard, we and others have shown the expression of gut-homing receptors, α4β7, on T_regs cultured in the presence of all-trans retinoic acid (ATRA) (Scotta et al., mauscript submitted). This may have important implications in the use of T_reg cell therapy in the context of inflammatory bowel disease. However, ensuring that T_regs express the relevant receptors and maintain their expression during the expansion process is challenging, as indicated by a recent study showing changes in the chemokine receptor expression of T_regs in vitro 99. In this study they showed that ex-vivo-cultured T_regs retained the expression of CCR7, but down-regulated CCR5 dramatically compared with freshly isolated T_regs.

Dose of T_regs and number of injections

Aside from the timing of injection and the site of injection, what is of paramount importance is to decide the dose of T_regs that is needed (recently reviewed in 100).

The trials to date (outlined below) of T_reg therapy in the context of bone marrow transplantation will inform us of the doses that are safe and tolerated in patients. However, the first trials of T_reg therapy in solid organ transplantation will need to start with a dose escalation study to assess the safety and tolerability of T_regs at various doses (minimum determined by the trials to date and the maximum number determined by the maximum number of T_regs reached during the in-vitro expansion process). In addition, whether polyclonal T_regs or antigen-specific T_regs are used will influence the dose. Of note, studies using antigen-specific T_regs showed that lower numbers were able to achieve the same functional efficacy as larger numbers of polyclonal T_regs 86,87. Finally, whether a single injection or multiple injections are required is a matter of debate and may be determined in a Phase II efficacy study, where patient outcomes should also be measured and an in-depth patient monitoring planned.

The use of molecular diagnostic tools can help to assess the increased expression of biomarkers of operational tolerance in patients receiving cellular therapy and whether these can be used as surrogate end-points of efficacy 101–103. The same approach can be used to define whether or not the patients have decreased expression of biomarkers of acute rejection 104,105. Furthermore, phenotypic analysis of patient PBMCs, using flow cytometric analysis, can determine whether or not the number of T_regs has increased or the composition of the T cell compartment has changed as a result of the intervention 106. Using the same analysis, the cytokine profile of the cells that have been phenotyped can be analysed to establish their plasticity. Finally, lymphocyte compartments can be monitored after specific interventions, as has been shown useful when associating expansion of lymphocyte subsets, in this case naive B cells, in peripheral blood of patients in whom better outcomes were noted 107.

First results of clinical trials

In spite of the potential concerns and controversies outlined with regard to T_reg isolation and expansion protocols and the optimal clinical protocol, clinical trials are under way to test the therapeutic potential of T_regs.

Beneficial effects of T_reg infusions on allograft survival were first reported in bone marrow transplantation models in which donor T_regs reduced the incidence of GVHD. The first human trial using T_reg cell therapy by Trzonkowski et al. 108 involved two patients. The first patient had chronic GVHD 2 years post-bone marrow transplantation. After receiving 0·1 × 10⁶/kg FACS purified ex-vivo-expanded T_regs from the donor, the symptoms subsided and the patient was withdrawn successfully from immunosuppression without evidence of recurrence. The second patient had acute GVHD at 1 month post-transplantation, which was treated with several infusions of expanded donor T_regs. Despite initial and transitory improvement, the disease progressed and resulted ultimately in the patient's death. This was the first report to show that adoptive transfer of T_regs is well tolerated and thus was a major breakthrough.

Results of a larger Phase I/II study were reported in which a total of 23 patients receiving umbilical cord blood (UCB) stem cell transplants were enrolled into a T_reg escalation trial 71. CD4⁺CD25^hi T_regs were isolated from a third-party UCB graft and expanded by anti-CD3/CD28-coated beads and recombinant IL-2 over a period of 18 days. Patients received expanded T_regs at doses ranging from 1 × 10⁵ to 30 × 10⁵/kg. Of note, the targeted T_reg dose was achieved only in 74% of cases. Compared with the 108 historical controls, there was a reduced incidence of grades II–IV acute GVHD (from 61 to 43%; P = 0·05), although the overall incidence of GVHD was not significantly different.

In a third trial (Phase I/II), conducted by Di Ianni et al. 109, 28 patients were enrolled who underwent haematopoietic stem cell transplantation for haematological malignancies. Patients received donor T_reg without ex-vivo expansion and donor conventional T cells (T_cons) without any other adjuvant immunosuppression. Different dose regimens were used, ranging from 5 × 10⁵/kg T_cons with 2 × 10⁶/kg T_regs to 2 × 10⁶/kg T_cons with 4 × 10⁶/kg T_regs. As two patients receiving the latter regimen developed acute GVHD, compared with none of the other patients, the authors concluded that a dose of 1 × 10⁶/kg T_cons with 2 × 10⁶/kg T_regs is safe. Moreover, patients receiving T_regs demonstrated accelerated immune reconstitution, reduced cytomegalovirus (CMV) reactivation and a lower incidence of tumour relapse and GVHD when compared to historical controls. However, it is also important to note the disappointing patient survival, with only 13 of the 26 patients surviving, but this may have been because of pre-existing fungal infections and the harsh conditioning regimens that were used.

With the results from stem cell-treated patients showing that T_reg therapy is well tolerated, it is now time to initiate trials in solid organ transplantation. In this regard, the ONE Study, a multicentre Phase I/II study funded by the European Union FP7 programme, will investigate the safety of infusing ex-vivo-expanded T_reg cells (among other regulatory cells) into kidney transplant recipients. Moreover, clinical trials to test the safety and tolerability of polyclonally expanded or donor alloantigen-specific T_reg cell therapy in combination with depletion of alloreactive T cells and short-term immunosuppression in liver transplant patients are currently being planned.

Conclusion

The first results of clinical trials applying T_regs in stem cell transplantation are very encouraging, and provide a basis for future trials in solid organ transplantation. Such trials should involve a small number of patients, aiming at evaluating the safety of increasing doses of T_regs. In addition, the clinical protocol for such trials should be based on a ‘T_reg-supportive’ immunosuppressive regimen, not only to protect against rejection, but also to create the tolerogenic milieu to maximize the potential efficacy of the exogenously administered T_regs. Recent progress in understanding T_reg biology and the development of experimental mouse models has highlighted how heterogeneous and plastic murine as well as human T_regs can be. This supports the importance of a careful design of purification and expansion protocols for generating T_regs for clinical application with release criteria set with the most current understanding of T_reg biology. Moreover, it is of paramount importance to ensure a comprehensive patient immune monitoring plan and the use of biomarkers that can predict the successful induction of immune tolerance, which would allow for the safe minimization or even withdrawal of immunosuppression.

Disclosure

The authors declare no conflicts of interest.