Introduction

Solid organ transplantation is a well-established treatment for patients with end-stage organ failure. Advances in surgical techniques and development of immunosuppressive drugs targeted at T-cell responses responsible for graft destruction have been key in improving short-term outcomes. However, there still remains the risk of chronic rejection and in the case of pancreatic or islet transplantation for T1D patients, therapeutic success is curtailed by recurrence of autoimmunity, despite immunosuppression that prevents rejection [1–6]. Moreover, chronic immunosuppression has significant toxicity and is associated with serious side effects, including increased susceptibility to infections. Therefore, there is a clear need to develop strategies for clinically applicable therapeutic protocols which are aimed at promoting tolerance induction that minimizes or eliminates long-term immunosuppression.

Immune tolerance is a state of unresponsiveness in the immune system. Self or natural tolerance is failure to attack the body’s own proteins or other antigens. If the immune system attacks its own cells or tissues, an autoimmune disease could develop. Induced tolerance is lack of an immune response to external antigens that has been generated purposefully by modulating the immune system. This type of tolerance is important in cases of transplantation where the patients’ immune cells “see” the allograft as foreign and mount an aggressive attack against it leading to graft rejection. Therapies in transplant setting are designed to target innate and/or adaptive immune responses that are thought to contribute to graft rejection. Another approach to immunotherapy would involve promoting tolerance mechanisms through enhanced regulation of adaptive T-cell responses. Central T-cell tolerance occurs in thymus where T cells develop and mature through a process of T-cell receptor (TCR) recognition of “self” antigens nestled in a major histocompatibility complex. Since high-affinity interaction could attack cells that display these self-peptides derived from body proteins, these T cells are eliminated through negative selection. Surviving thymic T cells leave and migrate throughout peripheral immune system (Fig. 1).

Fig. 1
figure 1

T-cell tolerance mechanisms. Central tolerance occurs in the thymus when high-affinity self-reactive T cells are eliminated. Low-affinity self-reactive T cells escape negative selection and migrate to the periphery, where they are controlled by several mechanisms, including regulatory T cells

Full size image

A central issue of modern medicine is how peripheral self-tolerance is achieved and preserved as increasing evidence has clearly demonstrated that potentially self-destructive T-cell clones can escape central tolerance and circulate in periphery of healthy individuals. There are several mechanisms for achieving peripheral tolerance, including anergy, negative selection occurring in the periphery, and control by regulatory T cells (Tregs) (Fig. 1). In this review, we will summarize the Treg population and their use as adoptive T-cell therapy for tolerance induction strategies. This review will also focus on thymus-derived, natural Tregs rather than induced regulatory T cells that are generated with TCR engagement and cytokines, since studies are demonstrating that natural Tregs are more stable population and potent suppressors of “self” and in preventing autoimmunity [7–9]. Treg-based immunotherapy could significantly impact development of new therapeutic protocols for transplant recipients and patients with autoimmunity based on administration of a T-cell population that could afford long-lasting protection from rejection and autoimmunity without the need for chronic immunosuppression. The first clinical trials with adoptive Treg therapy has occurred in graft-versus-host-disease (GVHD) prevention trials in bone marrow (BM) transplant recipients with promising results [10, 11]. However, Treg therapy could show promise in patients with autoimmunity, such as T1D, but pose more challenges due to the underlying autoimmune responses as well as in the context of allogeneic islet or pancreatic transplantation in these patients.

Adoptive T Regulatory Cell Therapy for Tolerance Induction

Adoptive immunity was first described by Billingham et al., in which studies showed transfer of primed immune cells could generate or “pass” immunity in recipient mice [12]. Since this time, several animal studies have shown that T-cell adoptive transfer of immunity in settings of cancer, infection, and transplantation [13–16]. T-cell adoptive therapy was then further extended in application by the use of IL-2 for ex vivo expansion of human T cells [17–20]. The idea of suppressor T cells was first suggested by two rodent studies from Gershon and Kondo [21, 22] that demonstrated the role of thymic lymphocytes in tolerance induction. A seminal study by Hall et al. demonstrated that CD4 T cells, specifically CD4+CD25+ cells, were capable of mediating transplantation tolerance [23••]. Furthermore, Sakaguchi’s group identified a crucial subset of CD4+ T cells that expressed the IL-2R α-chain (CD25) that functions in controlling peripheral tolerance and the development of autoimmune disease in mice and that CD4+CD25+ T cells from naive mice could prevent rejection of allogeneic skin grafts in nude mice receiving CD4+CD25− T cells [24•]. Since this time, there has been extensive study of CD4+CD25+ Tregs, which has further shown these cells are present in humans and function in all aspects of immune regulation [25, 26].

CD4+CD25+ Tregs naturally develop in thymus as a fully functional distinct CD4+ T-cell subset that migrates to periphery to actively suppress auto-reactive T cells that escape thymic negative selection. These CD4+CD25+ Tregs were later found to uniquely express the transcription factor Foxp3 [27–29] allowing more precise examination of their basic biology. Mutations or absence of Foxp3 leads to immune compartments devoid of CD4+CD25+ Tregs and mice die within 1 month of birth. The identification of mutations in Foxp3 gene in scurfy mice and in Immune dysregulation, Polyendocrinopathy, Enteropathy, X-linked (IPEX) syndrome in humans, both of which succumb to lethal autoimmune disease early in life, was critical in establishing an essential role of Tregs in the maintenance of peripheral self-tolerance. In mice, Foxp3 reduction or functional alterations results in spontaneous development of various organ-specific autoimmune diseases, including gastritis, thyroiditis, and diabetes [24•, 30, 31]. Adoptive transfer of CD4+CD25+ Tregs into Foxp3-deficient mice rescues disease development [27]. Importantly, these results following ablation demonstrated that natural Tregs are the dominant mechanism controlling self-tolerance and insufficiency of other mechanisms of peripheral tolerance in these mice.

Treg-dependent alloantigen tolerance has been induced in a variety of both in vitro and in vivo experimental models with active regulation critical for both inducing and maintaining immunological unresponsiveness to donor alloantigens [32–34]. Tolerance induction protocols in transplant setting require therapy to favors Treg-cell expansion and stimulation, while inhibiting alloreactive effector responses [32–39], thereby tipping balance towards regulation. Moreover, Tregs have not only been detected in recipient’s lymphoid tissues but also within the graft site [36]. The fact that Tregs can be found in multiple locations is probably critical for effective inhibition of an aggressive attack towards transplanted tissue. Furthermore, adoptive transfer of fully allogeneic Tregs into neonatal IL-2Rβ−/− mice prevents lethal autoimmunity associated with IL-2/IL-2R deficiency, but also confers tolerance to skin grafts bearing the MHC of donor Tregs [40]. Others have shown that donor or host ex vivo expanded Tregs together with allogeneic BM cells prevents GVHD and facilitates BM engraftment [41–43], but these studies did not examine persistence of donor Tregs or requirements for stable donor Treg engraftment. In the setting of allogeneic BM transplantation, substantial donor Treg engraftment initially may be sufficient to allow mixed chimerism to occur and reset immune system through central tolerance. However, in context of organ transplantation or autoimmunity, it may be necessary to maintain a population of donor Tregs, even perhaps antigen-specific Tregs, long term in order to establish and maintain a tolerant state.

Given the essential role Tregs play in self-tolerance [24•, 44, 45] and these cells suppress rejection of various allografts makes these cells highly attractive candidates for cell-based therapy for tolerance induction protocols in transplantation and autoimmune settings. Collectively, these experimental models suggest that Tregs might be used in non-toxic approaches for preventing GVHD, allograft rejection, and restoring self-tolerance. However, the clinical application of adoptive Treg therapy is hindered by low Treg frequency resulting in a limited number of cells to inhibit desired immune responses, lack of stable engraftment of donor Treg inoculums, and the need for antigen specificity of Tregs.

Adoptive Treg Therapy for Prevention of GVHD

There has been much success in Treg therapy to prevent GVHD [41–43], which has led to the first clinical trials with Treg immunotherapy in BM transplant recipients [10, 11, 46]. Since one obstacle in clinical application is low Treg frequency, many of these clinical trials have relied upon ex vivo expanded Tregs to generate large cell numbers to be transplanted. Furthermore, experimental models demonstrated the need for a high Treg/Teffector (Teff) ratio to promote therapeutic benefit [43, 47–49]. Several groups have pioneered in vitro Treg expansion protocols [32, 33] [10, 50–53] by polyclonal stimulation of autologous Tregs in the presence of high-dose IL-2. A confounding hurdle for ex vivo expansion of human Tregs is the observation that human CD4 T cells upon activation can transiently express Foxp3 and upregulate CD25 expression. Several studies have shown natural Tregs are largely found within CD4+CD25+CD127lo population in both mice and man [54–56].

In the first clinical trial with Treg adoptive therapy, two patients after hematopoietic stem cell transplantation (HSCT) with GVHD were given expanded CD4+CD25+CD127lo cells and showed these expanded Tregs were safe and may have therapeutic benefit [57]. In a larger phase I/II trial, Brunstein et al. utilized Tregs isolated from umbilical cord blood to lessen the possibility of expanding effector T-cell populations and found umbilical cord Tregs were safe and reduced grade I/II GVHD [10]. In another phase I/II trial, heavily conditioned hematological malignant patients following HSCT received unexpanded Tregs were found to enhance immune reconstitution, lower CMV reactivation incidence, and lower tumor relapse and GVHD [11]. This study showed immune regulation with human Tregs and safety profile in these high-risk patients. Although these GVHD trials showed great promise, in solid organ transplantation or autoimmune settings will pose more challenges due to the fact that unmanipulated recipients may not respond to Treg therapy as successfully because of the large presence of alloreactive or auto-reactive T cells, respectively, leading to greater Treg numbers to achieve Treg/Teff ratios that will favor a regulatory environment.

Clinical Trials in T1D and Solid Organ Transplantation

Because of the key role played by Tregs in self-tolerance, much effort is being devoted to developing Treg therapy for recent onset T1D [58, 59, 60••, 61]. Clinical trials with autologous, expanded Tregs are ongoing in T1D [60••, 61]. At present, trials are testing unselected, polyclonal Tregs, with awareness that expanded populations may include proinflammatory cells and may have limited lifespan in vivo as well as limited survival in non-lymphopenic hosts [52, 57]. A published phase 1 study showed that in vitro expanded, autologous Tregs were safe and tolerable in children with recent onset T1D, with evidence of improved fasting C-peptide and reduced insulin requirement at 4 months; therapeutic effects correlated with increased Tregs post-infusion, but persisted only for a short period of time. An extended follow-up confirmed limited persistence of expanded Tregs even after a second infusion [60••, 61, 62]. Data emerging from these recent Treg trials in T1D are indeed demonstrating limitations of current protocols that rely solely on infusion of expanded Tregs without any recipient manipulation, which strongly differs than in the clinical trials with HSCT patients.

With results in from Treg trials in HSCT and T1D patients demonstrating that adoptive Treg therapy is well tolerated, a trial in solid organ transplantation is now under way, the ONE study. This study is an international, multi-center phase I/II trial that will test safety and production feasibility of several regulatory cell populations, including ex vivo expanded Tregs, in living donor kidney transplant recipients. This trial aims at standardizing both experimental settings and endpoints and will likely allow collection of valuable data on the efficacy and safety of putative regulatory cell subsets for tolerance induction. In regards to Tregs, this trial will compare expanded Tregs from peripheral blood that is either polyclonal stimulated with anti-CD3 and anti-CD28 or alloantigen-driven with allogeneic antigen presenting cells (APC). Expansion protocol that utilizes stimulation with APCs, which enrich for allogeneic-specific Tregs rather than expanding the entire Treg pool, could allow for a more potent Treg population for adoptive therapy [63–66]. This type of expansion protocol is hindered by lack of sufficient time to expand recipients’ Tregs with donor APCs before transplant surgery. However, in the case of living-related kidney (or liver transplant) recipients, this approach is a more applicable strategy. Although the ONE trial is not designed to test the efficacy of the various regulatory populations, transplant patients will be immune monitored, which could provide valuable information on modulation of immune compartments following administration of regulatory cell populations and foster development of improved adoptive cell therapy for design of future clinical trials.

Key Requirements for Successful Adoptive Treg Immunotherapy

Although the development of ex vivo expansion Treg protocols have opened new avenues for adoptive Treg therapy in a clinical setting, there is still not much known about the in vivo environment necessary to support transferred donor Tregs long-term. Indeed, an often overlooked hurdle that needs to be considered for successful Treg therapies that avoid the use of chronic immune suppression is to understand the host environment. Crucial factors required for Treg generation, homeostasis, and function must be incorporated in design of clinical trials based on Treg adoptive therapy, whether these Tregs are expanded in vitro or infused after isolation for in vivo expansion.

Critical Barrier of Available Space and Minimizing Competition With Host Tregs

As described above, mutations or alterations in Foxp3 or IL-2/IL-2R interactions lead to immune compartments devoid of CD4+CD25+ Tregs and mice die within 1–2 months from birth. Importantly, adoptive transfer of wild-type Tregs into Foxp3-deficient mice prevents disease development in these models [73], and our own studies in IL-2Rβ−/− mice showed Treg adoptive transfer leads to prevention of autoimmunity and tolerance to skin allografts with life-long donor Treg engraftment and suppressor function. Unlike normal mice that have a constant production of thymic Tregs, a key feature of these mice is lack of continuous competition with endogenous Tregs because of absent thymic production and peripheral maintenance of functional Tregs resulting in a natural space for donor Tregs to engraft and persist. Moreover, adoptive Treg transfer into these mice occurred shortly after birth when the neonatal environment is initially lymphopenic followed by gradual increase in lymphocytes until homeostatic numbers are obtained as lymphocytes develop and migrate to periphery. Neonatal IL-2Rβ−/− model demonstrates the importance of available peripheral space and lack of competition to long-term, stable donor Treg engraftment. When a similar adoptive transfer of Tregs was performed in normal neonatal mice, there was 10-fold lower donor Treg engraftment compared to IL-2Rβ−/− mice [67••], indicating that competition with endogenous Treg production did not allow significant donor engraftment. Several investigators have shown positive therapeutic outcomes when Ag-specific NOD Tregs were adoptively transferred into NOD mice [68, 69]. However, these experimental approaches used recipient NOD mice that were devoid of either T or Treg cells, which results in natural space and lack of competition from endogenous Tregs, which likely promoted donor Treg engraftment, resembling the IL-2Rβ−/− and Foxp3−/− models.

In experimental transplantation models, successful adoptive Treg therapy was achieved in likewise immuno-compromised recipients, including irradiated, athymic, T-cell depleted, or Rag−/− hosts [70–74]. Although long-term donor Treg engraftment was not examined in these studies, it is likely that the recipient lymphopenic environment promoted donor Treg engraftment due to the reduced competition from host immune cells. For example, ablative conditioning severely depletes lymphocytes, including cells in the primary and secondary immune tissues, and lymphocyte reconstitution is exceedingly delayed, which recapitulates the in vivo environment in adoptive transfer in neonatal Foxp3−/− and IL-2Rβ−/− models. Collectively, these studies indicate that for successful Treg cell therapy, a critical obstacle that needs to be overcome is the competitive barrier created by the endogenous Tregs.

These observations have led to efforts in our laboratory to manipulate the Treg niche (“space”) and minimize competition with host Tregs in wild-type mice with the goal of improving adoptive Treg immunotherapy. This critical requirement was confirmed in common laboratory mouse strains in the absence of genetic manipulations, specifically in studies of BM conditioning in non-autoimmune C57BL/6 and Balb/c strains [75]. Infused Tregs engrafted, persisted long term, and contributed to tolerance induction to skin allografts. However, a possible concern in these studies was that infused donor Tregs will not maintain their Treg phenotype; it has been shown that adoptive transfer of Foxp3GFP+ Tregs into lymphopenic hosts results in a large number of cells to become GFP-negative [76–78]. However, the loss of Treg phenotype was dampened when Tregs were co-injected with CD4+GFP-negative cells and Foxp3 loss emerges in the absence of IL-2 [76]. Despite initial lymphopenia induced by our BM conditioning, donor-positive CD4 T cells maintain Foxp3 expression in both mouse strains [75]. These data together with immuno-compromised strains revealed the importance of available “space” within the Treg niche and minimizing competition at the time of Treg infusion for long-term donor Treg persistence. These studies also exemplify the success in HSCT models in both experimental rodent model and in clinical trials in the prevention of GVHD.

Antigen Specificity of Adoptively Transferred Tregs

Antigen has been shown to be another key requirement for Treg development and homeostasis. The ability of Tregs to prevent thyroiditis was lost when thyroid had been ablated, but rats retained the ability to prevent autoimmune diabetes [79], suggesting that the presence of antigen maintains a population of antigen-specific Tregs. Moreover, Tregs could be found within tolerant grafts, indicating that alloreactive Tregs could potentially be recruited to or even retained at sites of alloantigen expression and actively participate in tolerance induction and/or maintenance [80–83]. Other studies showed constant presence of alloantigen was required to inhibit allogeneic responses and induce tolerance to allogeneic BM cells [70, 71]. We also have shown a specific enrichment of donor Tregs to host alloantigen following adoptive transfer of allogeneic Tregs into IL-2Rβ−/− mice, indicating Treg adoptive transfer allowed biological selection of donor inoculums resulting in therapeutic Tregs capable of regulating autoimmunity and tolerance [40]. This dominant transplantation tolerance was towards the MHC antigens that the Tregs were initially selected for. Collectively, these data have important implications in autoimmune settings. For example, in fully diabetic NOD (or T1D patients), insulin-specific or other islet-specific Tregs may not engraft due to the loss of their cognate antigens. However, antigen-specific Tregs could work in conjunction with antigen therapies, in prediabetic or recent onset diabetes where adequate β-cell mass still exists, or in islet transplant settings. Importantly, in transplant setting, allografts will act as the source of cognate antigen. Alternatively, allogeneic Tregs from graft donors could be infused to potentially replenish these antigen-specific populations, but pose more of a challenge than recipients’ antigen-specific Tregs because of their immunogenicity. These types of strategies would involve adoptive transfer of antigen-specific Tregs using protocols that favor their long-term engraftment through in vivo antigen-driven selection and expansion for improved therapeutic efficacy. Furthermore, a major benefit from utilizing antigen-specific donor Tregs is that these Tregs will act only where antigen is present, thereby providing a local immunotherapy.

Studies showed that Tregs that are antigen-specific are more efficacious than polyclonal Tregs at regulating autoimmune diabetes and responses to alloantigen [68, 83–85]. NOD mice are notoriously resistant to tolerance induction protocols [86–88]. Nonetheless, islet-specific Tregs could have positive effects on diabetes prevention and diabetes remission in recent diabetic NOD mice. Islet-specific Tregs were isolated from BDC2.5 NOD transgenic mice and were expanded ex vivo in the presence of IL-2 and then adoptively transferred to immuno-compromised NOD mice, including Rag−/−, Scid, or TCRα−/− NOD recipients together with either splenocytes from diabetic NOD mice or activated BDC2.5 effector T cells. Diabetes prevention was achieved with very low Treg/Teff ratio (1:160) and as few as 50,000-expanded BDC2.5 Tregs [69]. However, when expanded BDC2.5 Tregs were given to 13-week-old wild-type NOD, diabetes was prevented with 30-fold more Tregs (1.5 × 106), while diabetes remission was achieved in only 36 % of recent diabetic NOD mice with two doses of 1.5 × 106-expanded BDC2.5 Tregs [89]. In another study by Bluestone’s group, 60 % diabetes remission rates were achieved when 107-expanded BDC2.5 Tregs were given to recent onset NOD mice [68]. These studies show that antigen-specific Tregs were efficacious compared to expanded NOD Tregs; however, success was achieved with low Treg numbers only when recipient mice had available “space” and lack of competition with host Tregs, indicating that similar success in autoimmune or transplant patients will require immunomodulation to recreate this in vivo environment to support their engraftment and therapeutic benefit.

In transplant setting, expansion of recipients’ polyclonal Tregs with allogeneic APC leads to alloantigen-specific Treg enrichment allowing more potent Treg population for immunotherapy [63–66]. Sagoo et al. developed a method based on expression of CD69 and CD71 following stimulation of CD4+CD25hiCD127lo T cells with allogeneic dendritic cells (DC). These activated Tregs were readily expanded in vitro in the presence of allogeneic DC and IL-2, and in a humanized mouse model these enriched Tregs could significantly reduce allogenic immune-mediated skin damage compared to polyclonal activated Tregs [64]. In another study, Veerapathan et al. showed CD4+CD25+CD127lo T cells that were stimulated with allogeneic DC in the presence of IL-2, IL-15, and rapamycin were enriched and occurred when alloantigen was presented directly or indirectly with autologous DC pulsed with allogeneic fibroblast lysates. Either population was more potent suppressors in vitro than polyclonal expanded or non-expanded Tregs [66]. This group also demonstrated expansion of Tregs against minor major histocompatibility antigens [65]. These studies show promise that alloreactive Tregs can be selected and expanded and are more potent than polyclonal Tregs in vitro, but failed to examine graft outcomes. Enriched allogeneic-specific Tregs will be tested in the ONE clinical trial, and results will allow for development of improved clinical trials such that enriched donor-reactive Tregs given after transplantation could lessen immunosuppression dosing and duration. Expanding donor-specific Tregs would be possible when living related donors are used, but expansion may be cumbersome to implement from cadaveric organ donors. If Treg expansion protocols are shortened or if Tregs could be administered days after transplantation, these would allow sufficient time for ex vivo Treg expansion and could overcome this hurdle for their therapeutic application in these types of transplant patients.

IL-2: Critical Factor for Treg Cell Development and Homeostasis

The expression of CD25 was the initial marker used to identify a population of Tregs. Several studies, including our own, have proven that IL-2/IL-2R (IL-2 receptor) interaction represents an essential step in production and supports Treg homeostasis in steady state [67••, 90–97] (Fig. 2). The role of IL-2 in Treg generation and maintenance has been extensively reviewed (see [98, 99]). Here we will focus on the role of IL-2 in the context of Treg therapies. If significant long-term donor Treg engraftment will require immunomodulation prior to Treg infusion, it may be possible that other STAT5-dependent cytokines, such as IL-7 or IL-15, may act as homeostatic factor rather than IL-2 during these lymphopenic states. However, residual Treg compartment following ablation was also highly dependent on IL-2 and IL-7 acted more as a survival factor [100]. Moreover, IL-2 was also critical for the growth and survival of infused Tregs following conditioning [75]. Despite low IL-2 in conditioned mice, levels are sufficient to drive substantial proliferation Treg compartment. This is likely due to Treg homeostasis requiring low IL-2 signaling [101] and their expression of high affinity IL-2Rs over other lymphocytes allows competitive advantage for IL-2 utilization. Impaired Treg function in NOD mice derives from low IL-2 levels in pancreas; proper IL-2 treatment restored Treg function and effectively reverses diabetes at onset in a fraction of mice by augmenting the Treg/Teff ratio [102]. Furthermore, enhanced β-cell destruction occurs by Th1 and Th17 CD4+ T subsets in recent diabetic mice, but low-dose IL-2 administration downregulated production of these cytokines [102]. In contrast, high-dose IL-2 triggered disease likely by promoting autoreactive T cells.

Fig. 2
figure 2

Role of IL-2 for Treg development and homeostasis. The essential role for IL-2 resides at the earliest stages of the production of Tregs in thymus. The IL-2/IL-2R interaction is active in thymus at the earliest stage of Treg development to promote Treg expansion and to upregulate Foxp3 and CD25 to normal levels. These Tregs exit the thymus and home to neonatal lymph nodes where IL-2 acts as a potent growth factor promoting substantial Treg expansion. In the adult peripheral compartment, IL-2 remains the dominant mechanism for Treg homeostasis, but mature Tregs can be maintained low-dose IL-2. In lymphopenic conditions, IL-2 is still the dominant cytokine controlling Treg homeostasis and maintaining high expression of CD25 and Foxp3. IL-7 acts as a survival factor to maintain Tregs in the periphery, whether during the steady state, lymphopenia, or following Treg infusion following ablation

Full size image

IL-2 continues to be used in cancer immunotherapy at high doses to boost tumor responses, but has had limited use in autoimmune setting due to the potential risks of stimulating effector T cells (for review, see [13, 103]). In phase I/II pilot study combining DCs and IL-2 in metastatic renal cancer patients, a specific anti-tumor response was detected, but a transient and massive increase of circulating Tregs was also observed [104]. A phase 1 clinical trial in recent onset T1D with a combination therapy of IL-2 and rapamycin was recently completed [105]. Although there was expansion of Tregs, there was proliferation of Teff and NK cells and a worsening of metabolic function. It appears that the IL-2 dose was not sufficiently low to achieve only specific effects on Tregs. In an autoimmune-related HCV-induced vasculitis clinical trial with low-dose IL-2, it showed specific Treg activation without stimulating other T cells and showed a good safety profile. Importantly, low-dose IL-2 results in increased Treg/Teff ratio favoring regulation, downregulates inflammation, and failed to lead to viral reactivation in these patients [106]. This group also recently reported that phase I/II study in T1D patients with low-dose IL-2 was well tolerated and IL-2 induced a dose-dependent increase in proportion of Tregs [107]. These clinical trials with IL-2 have important implications in developing Treg immunotherapy for the clinical setting in that the proper low-dose IL-2 combined with Treg infusion following immunomodulation may allow for far fewer numbers of donor Tregs for positive therapeutic outcomes and dampen the inflammatory environment in solid organ transplantation and in autoimmunity.

CD8 T Regulatory Cells: Potential Cellular Source for Adoptive Treg Therapy

CD4+CD25+Foxp3+ cells have been the main cell population to date that has been tested in clinical trials. As mentioned above, the ONE trial will examine several other regulatory populations to be potentially used for therapeutic benefit. However, Harvey Cantor’s group have identified an IL-15-dependent, Qa-1-restricted CD8 T cell with regulatory properties that has been shown to be essential in maintenance of self-tolerance and prevention of autoimmune disease in mice [108–111]. These cells can be identified and purified on the expression of the surface markers CD44+CD122+Ly49+. These cells could also be expanded ex vivo driven by peptide and infusion of these peptide-specific CD8 Tregs strongly inhibits collagen-induced arthritis mouse model by eliminating pathogenic T follicular helper and Th17 cells [111]. This group has recently identified the human counterpart to these murine CD8 Tregs. These human CD8 Tregs express KIR2DL2/3 or KIRDL1 along with expression of CD44 and CD122, and IL-15 promotes their activation and proliferation [112]. These cells could represent another regulatory population that could be considered in future adoptive Treg therapy given that these cells can be purified, expanded ex vivo, and can be antigen specific.

Concluding Remarks

Promoting immune regulation through enhancing Tregs offers the hope to suppress many unwanted immune responses in transplantation to prevent graft rejection, as well as a plethora of autoimmune diseases and GVHD. However, results in animal models and in recent clinical trials with adoptive Treg therapy are highlighting the need to consider the in vivo environment to support their engraftment, persistence, and function of infused Tregs for the future design of clinic trials in solid organ transplantation. Success in both experimental models and in clinical trials required immunomodulation to recapitulate a supportive Treg environment. Treg immunotherapy in autoimmune settings will pose more challenges given chronic inflammation could create a non-supportive environment for Tregs. However, the use of low-dose IL-2 could possible overcome the inflammatory environment and foster regulation over effector responses and promote a supportive environment for stable engraftment of infused donor Tregs. The transplantation setting could provide an adequate source of cognate antigen to drive antigen-specific regulation through selection and expansion and in context of autoimmunity where self-antigens are lost from immune destruction. Collectively, successful Treg immunotherapy will likely require combinational approaches in which Treg infusion occurs following immunomodulation to make space in recipient’s Treg compartment and to minimize competition with host Tregs accompanied with low-dose IL-2 to provide homeostatic support for Treg compartment for long-term persistence of donor Tregs (Fig. 3). However, introduction of immunomodulation could potentially have unwanted side effects that would need to be evaluated for combinational approaches. For example, the use of anti-thymoglobulin (ATG) in induction protocols leads to dramatic lymphocyte depletion and is associated with Treg expansion in vivo favoring a shift in Treg/Teff ratio. However, CD4 and CD8 memory T cells were resistant to ATG depletion and were found to expand following treatment [113, 114]. Recently, Tregs have been shown to play a critical role in natural killer (NK) cell homeostasis, activation, and function, predominantly by controlling the availability of IL-2 in the microenvironment [115–117]. In Treg absence, NKs produce abundant IFNγ and contribute to diabetic lesion in NOD mice and NKs were found to be initiators of autoimmune response by promoting CD4 T cells. In conclusion, future studies aimed at directly examining these issues in the context of clinically relevant approaches which could afford therapeutic synergy and considerably reduce the need for large numbers of Tregs for therapy will provide the framework for development of novel strategies for tolerance induction.

Fig. 3
figure 3

In vivo environment for Treg Immunotherapy. To recapitulate the in vivo environment to promote stable Treg engraftment will require immunomodulation prior to Treg infusion. This immunomodulation needs to create initial space and minimize competition with rebounding host Tregs to promote donor Tregs to engraft and persist. The addition of low-dose IL-2 for homeostatic support for the Treg compartment, which will be composed of both recipient and donor Tregs, will foster Treg numbers, but may also play a role in stabilizing Treg signature and downregulate inflammation

Full size image